Ron Vanderkley – Robohub

ep.294: Autonomous Bricklaying by FBR, with Mark Pivac

Ron Vanderkley — Wed, 18 Sep 2019 17:15:07 +0000

In this episode, Ron Vanderkley interviews Mark Pivac, Chief Technical Officer and co-founder of FBR (formerly Fastbrick Robotics) about the world’s first end-to-end autonomous bricklaying robot, ‘Hadrian X’. Three years after his first interview, we catch up with Pivac to see how FBR has expanded its operation and chat about their latest commercial prototype, ‘Hadrian X’, as well as the future of the robotic construction industry.

Mark Pivac is the primary inventor of FBR’s automated bricklaying technology. He is an aeronautical and mechanical engineer with over 25 years’ experience working on the development of high technology equipment ranging from lightweight aircraft to heavy off-road equipment. He has 20 years’ experience working with pro/engineer 3D CAD software as well as high-level mathematics, including matrix mathematics, robot transformations and vector mathematics for machine motion. Mark has also worked extensively with design, commissioning and fault finding on servo controlled motion systems achieving very high dynamic performance.

Links

ep.226: Toru Robots, with Dr. Moritz Tenorth

Ron Vanderkley — Sat, 21 Jan 2017 23:37:49 +0000

In this episode, Ron Vanderkley spoke to Dr Moritz Tenorth, head of software development at Magazino, a Munich-based startup developing mobile pick-and-place robots for item-specific logistics. They discussed his work on the Toru robot and what it means to the warehouse industry today and in the future.

Dr Moritz Tenorth

Dr. Moritz Tenorth

After obtaining his PhD in robotics from TU Munich Dr Moritz Tenorth spent several years as a post-doc and freelance robotics consultant, with stays at the CMU Robotics Institute and the ATR labs in Japan. His research focuses on knowledge representation methods that can help autonomous robots make smarter decisions, and on methods for task coordination in robotics.

Links

ep.217: LunaRoo, with Jürgen Leitner

Ron Vanderkley — Sat, 17 Sep 2016 19:37:07 +0000

In this episode, Ron Vanderkley interviews Jürgen “Juxi” Leitner, a researcher at the ARC Centre of Excellence in Robots Vision in the Queensland University of Technology in Australia. Leitner speaks about a system being developed for the Google Lunar XPrize, called LunaRoo.

The Google Lunar challenge calls for privately funded spaceflight teams to land a robotic spacecraft on the Moon, travel 500 meters, and transmit back high-definition video and images. LunaRoo is a hopping robot designed to leap up to 20 meters in the Moon’s lower gravity. While in the air (well, not really), LunaRoo will use computer vision to track the movement of a rover, that is trying to drive 500 meters, navigate the Moon’s surface.

Below is a video showing some prototypes of LunaRoo.

Jürgen Leitner

Jürgen “Juxi“ Leitner is a researcher at the ARC Centre of Excellence in Robotic Vision. Before coming to Australia, he worked and studied at various research institutes in Europe and Japan. From 2010-2014, he worked in a robotics lab at the Dalle Molle Institute for AI (IDSIA) in Lugano, Switzerland. Prior to IDSIA, he worked at the European Space Agency‘s Advanced Concepts Team for a year.

Links:

ep.205: Hadrian Bricklaying Robot, with Mark Pivac

Ron Vanderkley — Sat, 02 Apr 2016 21:13:08 +0000

Transcript below.

In this episode, Ron Vanderkley speaks with Mark Pivoc from FastBrick about Hadrian the bricklaying robot. Fastbrick is an Australian robotics firm aiming to disrupt the local bricklaying market with a machine it says will be able to build a four-bedroom house in two days, without any human interaction.

A builder or architect will supply a 3D model, which is then converted into machine code that uses proprietary software to calculate the location of every brick in the building. Laser-scanning technology surveys the foundation once a concrete slab has been laid, and then loads and lays bricks using a 28m articulated telescopic boom. Adhesive is applied to the brick via the robotic laying head, and a laser alignment system ensures an accuracy of within half a millimeter.

Mark Pivac
Mark Pivac is the primary inventor of Fastbrick’s automated bricklaying technology. He is an aeronautical and mechanical engineer with over 25 years of experience working on the development of high technology equipment ranging from lightweight aircraft to heavy off road equipment. Mark has 20 years of experience with pro/engineer 3D CAD Software. He also has high level mathematical experience including matrix mathematics, robot transformations and vector mathematics for machine motion. In addition, he has extensive design, commissioning and fault finding experience on servo controlled motion systems achieving very high dynamic performance.

Links:

Transcript
This transcript has been edited for clarity.

Ron Vanderkley: Welcome to Robots Podcast. Could you introduce yourself to our podcast listeners?

Mark Pivoc: I’m Mark Pivac. I’m the Chief Technical Officer of FastBrick Robotics. We’re developing the world’s first fully automated end-to-end bricklaying robot.

Ron Vanderkley: Why a bricklaying robot?

Mark Pivoc: Well, people have been laying bricks for over 6,000 years and have been trying to automate the process and the intellectual property prior goes back to the times of the Industrial Revolution. The big issues at the moment for automation really are tackling anything that’s dull, dirty or dangerous. Bricklaying actually ticks all three of those. Although it’s not immediately dangerous, over a long period of time there’s a lot of repetitive strain injuries. It’s a job which in Australia is out in the sun, so there’s quite a high skin cancer risk.

Any jobs which are dull and dirty and repetitive, and bricklaying certainly is, it’s very difficult to attract young people to take up the trade. The bricklaying organizations and the builders are having a lot of trouble training people. Even once they do attract a few people to start up the bricklaying apprenticeships, they have over a 50% dropout rate. Really, it’s a supply side driven issue in that there’s not enough bricklayers, particularly in Australia, but we also know that it’s throughout the world. In lower labor cost countries, there may be a pool of labor which is willing to do the bricklaying but they have issues with getting the appropriate skill and the training of those people to become bricklayers.

Although bricklaying is a repetitive task, it’s actually quite a craftsman task because bricklayers don’t just put bricks down. A lot of the other work they do involves setting out the actual size of the house, making sure that it’s the right dimensions and square and that the windows and doors end up in the right place. It’s actually quite a skilled job. Although somebody can be trained up in an few days or weeks to become a productive bricklayer that’s just putting bricks down in a straight line or on a straight wall, it actually takes a full apprenticeship and years of training to really understand the intricacies of building an entire house from scratch from bricks to the modern standards.

Of course, 6,000 years ago people didn’t really care, but these days there’s a whole bunch of different standards and requirements which vary depending on the location of the house, whether it’s in a cyclonic region and needs special roof tie-down straps put in or whether it’s close to a coastal environment and needs different mortar specification or in a bush fire area where the bricklaying needs to meet certain fire requirements or in commercial buildings, again, where you get certain fire requirements.

Of course, the modern requirements for insulation and thermal performance of buildings is meaning that the bricklaying process is becoming more complicated, with bigger bricks, pre-insulated bricks, or insulation added between the bricks. You add up all these various complications, and it really is pretty much a no-brainer that bricks are going to be used into the future of building houses or commercial buildings. It really needs to become an automated process, and that’s why we’ve tackled it.

Ron Vanderkley: Let’s talk about your product. Can get you to describe your robot to our listeners?

Mark Pivoc: Our prototype robot is a 25 ton machine which can reach out 28 meters. The idea of it is you deliver the machine to site, you park it once at the building site, and then the boom is big enough to reach over the entire building site. If you would like to picture how this thing works in reality, the building site will be pre-prepared and the concrete slab laid, and then the bricks delivered to site. We turn up with our machine. A telehandler is used to load the bricks onto the machine. The machine can take three packs of bricks.

There’s automated de-hacking robots which take a row of bricks off each pack and place them onto a shuttle, which then transports the bricks either to a saw or a router or a transfer robot, which then places them onto a chain conveyor where the bricks are gripped by little screw grippers. They are transported then out to a laying head, which is very a dynamic spherical geometry robot that has linear motors and direct drive top motors on it and also some harmonic drive motors. That laying head, the position is monitored in six degrees of freedom. Three linear XYZ coordinates and also the orientation by laser tracker and a smart track sensor.

Those positions are read in a very high speed, some are in the order of 100 Hertz to 1,000 Hertz, into our CNC control system which calculates how much the boom has moved from where it actually should be — that could be due to slight mechanical inaccuracies, wind, or just the dynamic motion of it moving around. We know precisely to fractions of a millimeter and fractions of a degree how that end of the boom is positioned and orientated. We work out where the gripper should be and where it actually is and apply a little compensation there so that the robotic arm of the laying head plucks the bricks off the end of the boom and places them.

Just prior to that, a layer of adhesive is applied to the bricks. We use adhesive for a number of reasons, as opposed to mortar. We have used mortar in the past and we can handle it, but there are a number of issues with mortar. The main one is that it has quite high thermal conductivity relative to the bricks. The modern construction methods are shifting towards a foaming adhesive, which is quite insulating and has less of a thermal bridging effect compared to mortar. We apply just a bed of adhesive, the horizontal joints of the brick.

We don’t do the vertical joints because our initial targeting is for the render construction market. When you apply the render that goes into the vertical joints and it helps the render key on and also provides that bonding on the vertical, what’s called the vertical perp joints. We are in the long run aiming to lay face brick, but we do intend to do that with, again, either a thin bed mortar or an adhesive. The bed will finish on the edges of the bricks to give that jointed effect. We could use mortar, but the modern thermal requirements are moving away from that. Also, there’s quite a big cleanup involved.

The adhesives we use don’t set until they’re exposed to either the oxygen or the moisture in the air. When they’re inside the machine there’s no cleanup problems in the pipes and so on, whereas mortar and thin bed mortars are a cement-based product, which is a hydration reaction. As soon as you mix it with the water they start to set. You’ve got a limited time frame, and then of course you have to clean it out. With up to 16 meters of hose on a machine, that becomes quite an issue in itself. Really, our whole approach has been to look at all the labor and time-consuming processes of either bricklaying or operating a machine and trying to design out those time-consuming and expensive human aspects from the machine.

Ron Vanderkley: On a high level, what type of competition do you have? I’ve heard of robots like the one in the U.S., SAM. What sets your product apart?

Mark Pivoc: Our main difference is that we handle the bricklaying process from end to end. From the moment the bricks are delivered to the building site through to the time of the finished house or structure, we’ve tackled it with an automated approach. Our competitors, like SAM in the U.S., they require manual feeding of the bricks to the robot quite close to where the bricks are being laid. In its current version, SAM can only build one wall at a time. It needs to be taught where the windows and doors are along that wall. It also needs a mason, a trained bricklaying person, to be there to help it out with certain things and do the setup.

Our process is very automated right from the design phase. We take it through a SOLIDWORKS model and we’ve written special API interfaces to that to calculate all the bricks and cuts should happens. Even the generation of our g-code program is quite automated. We handle the bricks automatically from the moment they’re placed on the machine, which of course replacing packs of bricks being done by a forklift or a telehandler. From there right through to the laid brick it’s untouched by human hand.

Ron Vanderkley: When we talked about describing your machine, you mentioned it didn’t have hydraulics. It’s just driven by electric motors and servo.

Mark Pivoc: The entire machine is servo electric. Our first prototype has quite a lot of pneumatic components on it, but our next prototype is moving to be almost entirely electric. The only pneumatics on it really will be used for cleaning and contamination control.

Ron Vanderkley: We talked about the software, and that you’re using a SOLIDWORKS model and you’re creating CNC code directly from that to drive it.

Mark Pivoc: That’s been one of the major parts of our project, has been figuring out a good way to come up with the program for bricklaying. We’ve looked at a lot of different ways of doing it. Although it’s a similar process to 3D printing in that we build up layer by layer, we have to consider that each brick is finite volume and fixed size. Whereas with 3D printing, you’re dealing with something which is a lot more flexible in terms of it’s a much smaller voxel, or 3D pixel if you want to look at it that way. We’re building up a structure out of fairly big voxels, or 3D pixels, which are not cubic. They’re a rectangular prism. We have to deal with that, and we have to manage a lot of complication and issues to do with the way they’re cut.

We can’t just cut off, say, a 10 millimeter thick sliver of brick. How do we handle that throughout the rest of the machine? It’s really, like most computer programming, you can bang out the basic guts of how the program works in maybe a day or two and then you spend the next five years perfecting it to handle all the intricacies and complications. Certainly, that’s been a big part of our process of development, has been dealing with all those intricacies, the cut bricks and corners and getting the bricks to mesh properly and figuring out practical ways or operators to be able to program this thing in a way that is faster than a bricklayer in building the actual house.

Ron Vanderkley: What have been the biggest problems you’ve overcome?

Mark Pivoc: The biggest problem for us has always been, until quite recently, the financial problem. We know that there’s a fantastic market out there. The annual bricklaying market worldwide is just absolutely massive. I got extremely interested in this project when we came up with the concept for an automated bricklaying machine. I really had no idea how many bricks were being used. All I knew what that bricklaying was in the news because there weren’t enough bricklayers available and the price of laying a brick had doubled in 12 months, which is usually an indication that there’s something going on in the market, either a lack of supply or a massive increase in demand.

I jumped on the Bureau of Statistics website and saw that there are over 2 billion bricks a year being made in Australia. I figured that at a dollar a brick to lay them, it was a pretty good market and one worth tackling. We approached the design of the machine without too much concern about how much it would cost because we knew there was a good market at the endpoint. Like all things in automation, as your production numbers go up, the costs come down. We figured that we’d be able to get the costs under control later on. As you can appreciate, it’s a pretty massive automation task that covers a lot of different disciplines and is basically trying to build an entire production line or production plant on something that can be moved from site to site, operated outside in pretty much all sort of practical weather.

From a technical point of view, most engineers know how to deal with those sorts of issues, but it’s getting the funding to sort that out. In Australia for a long time there’s been a reluctance to fund either manufacturing or technology based industries. We’ve had a very busy mining environment. As a consequence, the financial people have been more interested in mining. Unfortunately for Australia as a whole, but fortunately for the tech sector, there’s been a big downturn in mining and commodity prices and the financial capital markets are looking for technology and manufacturing and apps and all those sorts of things. It’s been a case of working at a slow rate for eight years to get the timing right so that it matches up with the financial markets. We’re well funded now, and now we can get in place a lot of technical solutions that just weren’t financially possible before and bring our product to market.

Ron Vanderkley: What comes after this? Currently, you’ve got a prototype. You’re going to be going to production, is it next year?

Mark Pivoc: We have a series of prototypes and production machines coming up. Our first prototype, which we call the 105, is really a technology demonstrator to prove that it’s possible. We know that there’s a number of things with that 105 machine which mean that it’s not likely to be used building houses on site. It would be suitable for some other application, such as a freeway sound wall and things like that, but it’s not ideally suited to the Australian residential market in its current form. We’re pretty well progressed on our next prototype, which we call the 109, and we see that as being the pre-production prototype, which we’ll start building early in 2016. Provided that one passes all of its tests and meets its expectations, then that will become the basis for our production machine.

Ron Vanderkley: Is there anything else planned? We spoke about the possibility of rendering machines. Are there other attachments?

Mark Pivoc: There’s a lot of different possibilities for large scale robotic construction and automation. The key technologies that we’ve developed are what we call course find positioning, so that we can approximately position over a very wide area and then measure very precisely that position and have a fine manipulator, which can then do quite precise work in a much smaller volume. Because it can be moved throughout a big volume, it means we can do tasks which would normally only be possible to automate in a factory with fixed robots. We can now tackle those tasks outside or in a much bigger area. There’s a whole range of different possibilities, ranging from riveting and automated fastening, automated welding, sandblasting, painting, all sorts of kinds of processes, rendering. If you can imagine that there’s a market big enough out there to justify the development costs, then really the world’s your oyster.

Ron Vanderkley: What are the job prospects in robotics in Australia? Are they getting better?

Mark Pivoc: I’m probably the wrong person to talk to. I’m more of a cup is half full guy than cup is half empty. To give you an example, I’ve got three mechatronics graduates who’ve just finished their course. They’re now employed with us here. From our point of view, we think there’s fantastic employment prospects for bright young engineers. Australia as a whole I think has still got a way to go, but it’s pretty widely recognized that there’s not enough engineers around the place. Certainly at the moment in WA there’s probably a few more engineers than the market wants at the moment, but give it a year or two and that’ll probably change.

A lot of that is WA is driven by the resources sector, and that follows through to the manufacturing sector. A lot of Australian manufacturing has disappeared over the last few years because of the high exchange rates, meaning that we’ve become fairly expensive in Australia. With the dollar sitting at about 72 cents as opposed to $1.10, we’ve just become 30% cheaper than we were a couple of years ago. I don’t know too many industries that can become 30% cheaper without doing anything. I think what’s probably a big issue for Australia is that we all understand it’s a global economy and we have to have a floating exchange rate, but floating to the degree that it does in Australia where in my working career I’ve seen the dollar as low as 52 cents and as high as $1.10.

Any long-term planning for industries which have to have presence for 20 or 30 years to become efficient, car manufacturing, there’s such a massive amount of infrastructure, training and technology and associated first, second, third, fourth tier suppliers that feed that production line. You can’t just set up a car industry or a robotics automation industry overnight. It takes years to develop the technology and the expertise to do that. Probably in Australia one of our biggest issues is the fluctuating exchange rate.

We’re going through the same thing now looking at where we’re going to produce these machines. The biggest factor of uncertainty in our production planning is what’s our exchange rate going to be in three years’ time? Nobody can crystal ball gaze that. Is it going to be 75 cents? Is it going to be $1.10? Is it going to be 52 cents? Is it going to be $2.00? Who knows? Which makes it extremely difficult to plan. Yeah, we’ve got some issues in Australia that we have to figure out some way to deal with.

Ron Vanderkley: On behalf of the podcast, I’d like to thank you, Mark, for your time. We’d like to keep in touch and see where your product is going in the future. Thank you.

Mark Pivoc: Thanks, Ron. It’s been a pleasure.

ep.196: Marine Robotics Systems, with Stefan Williams

Ron Vanderkley — Mon, 30 Nov 2015 09:57:00 +0000

Transcript below.

In this episode, Ron Vanderkley speaks with Stefan Williams of the University of Sydney’s Australian Centre for Field Robotics, Marine Systems Group. They discuss the future of Autonomous Underwater Vehicles (AUVs), and a recent expedition where they used multi-session SLAM to map the famous Antikythera Shipwreck (circa 60-80 B.C.), one of the richest ancient wrecks ever discovered. It is located under 55m of water on the NE coast of the island of Antikythera. The site is famous for the first Analog Computer known as the Antikythera Mechanism, a geared device designed to calculate and display celestial information, including phases of the sun and a luni-solar calendar.

The ACFR leads Australia’s Integrated Marine Observing System (IMOS) AUV Facility. IMOS is a nationally coordinated program designed to establish and maintain the research infrastructure required to support Australia’s marine science research. The IMOS AUV facility generates physical and biological observations of benthic variables that cannot be cost-effectively obtained by other means.

Stefan Williams
Stefan Williams is part of the Faculty of Engineering & Information Technologies, School of Aerospace, Mechanical and Mechatronic Engineering, University of Sydney’s Australian Centre for Field Robotics Marine Systems Group.

Links:

Transcript

Ron Vanderkley: Stefan, can I get you to introduce yourself to our podcast listeners?

Stefan Williams: I’m a professor of Marine Robotics at the University of Sydney’s Australian Center for Field Robotics. We undertake both fundamental and applied research in the area of marine robotics and the development of underwater systems.

Ron Vanderkley: What is the main goal of the Marine Group?

Stefan Williams: Our focus over the last decade or so has been to develop tools and techniques for better understanding marine environments. We work in a variety of different application areas, from ecology to archeology and geoscience, and we work in the engineering research areas that underpin these applications – such as vehicle navigation, control planning, machine learning, and computer vision – in order to provide better understanding and information to our end users about the data we are collecting.

Ron Vanderkley: When you talk about autonomous marine surveying systems, what kind of capabilities are we talking about today? Is this a mix of manual and autonomous techniques? Where are the AUVs?

Stefan Williams: The main work that we do is on autonomous underwater vehicles. These are vehicles that are disconnected from the surface. We program them with a particular mission that we’d like them to execute, which might be a big long transect or a particular depth profile, or it might include very detailed studies over a particular feature of interest. Then a vehicle goes down and executes that mission. We typically stay in communication with the vehicle using acoustic modems; the vehicle tells us what it’s doing, it gives us feedback on the data that it’s collecting, and we have some supervisory control to tell it “no, you’re in the wrong spot, please come up” or “there’s some issue with the mission so abort” for example. We also have some ability to redirect the missions while they’re underway.

The data is collected onboard the vehicles. Once we recover them we download the data and build detailed 3D models of these marine habitats primarily using vision data, but we also collect multi-beam sonar, so we have acoustic data as well if we’re looking to map broader areas of the sea floor. That’s the core technology that has been developed around these autonomous platforms, but we’ve also built systems for both shallow water surveys using divers, and also very deep-water exploration using remotely operated vehicles. We also have platforms and sensor packages that we can use in other ways beyond just the autonomous platforms.

Ron Vanderkley: You spoke about the acoustic modems. What level of data can you get out of that? Is it to the level of telemetrics? Can you actually sift some data instead of just downloading it at the end of the mission?

Stefan Williams: Most of the data that we get back at the moment is supervisory in nature. The robot will tell us where it is, how deep it is, how far it is off the bottom, and the amount of data or number of images that it has collected. This lets us monitor the progress of the mission and make sure that things are on track. That modem is also piggybacked on positioning information, so it tells us where the vehicle is as an independent measurement from the ship. We then send both the ship position and the observations of the robot’s position relative to the ship down to the vehicle, and it updates its estimate of where it is using that information. We’re able to correct any drift in the navigation estimate using those external observations.

We’ve just taken on a new positioning and modem package in the past year or so that is capable of higher baud rates. In theory, we can get thumbnails of images. I’ve seen some work where they’re actually using this same sort of acoustic modem to do low-bandwidth video across an acoustic length, but that’s a capability we still have to develop.

Ron Vanderkley: I believe that your group is working on some novel enhancement to acoustic communication from the operator to the robot. Is that to increase the baud rate?

Stefan Williams: We’ve just been awarded an Australian Research Council Linkage Grant, which encourages the research community to engage with interested partners. This is a partnership with Tellus Australia – they do a variety of different developments in the marine space, particularly focused on defense applications. The project is really about managing the communication channels: it’s less about developing new acoustic communications equipment than it is about what information is best to send across the acoustic communication channel so that the operator can most effectively monitor the progress of the mission. In an information theoretic sense, we want to decide what information is best from the point of view of giving the operator an idea of what’s happening, and then allowing that operator to make decisions and query for further information from the vehicle. For instance, it might be used to refine and estimate where particular targets are in an environment, or to request more detailed high-resolution components of a particular area that might be of interest to the operator.

Ron Vanderkley: I also was aware of a multi-session SLAM-based AUV mapping system, and there was an example of a wreck that you worked on. What did that entail?

Stefan Williams: We’ve been involved over the past year in projects together with the Woods Hole Oceanographic Institute and the Hellenic Ministry of Culture and Sport (and a few other organizations) in mapping and excavating a 1st century BC shipwreck off the coast of the Greek island of Antikythera. This wreck was first discovered in 1900 and excavated around that time, yielding a whole trove of statues and other ancient artifacts – including something called the Antikythera Mechanism, which is one of the first examples of what people consider to be an analog computer. It was a geared mechanism that was used for predicting celestial motion over a fairly extended period of time to predict where the sun was and the phases of the moon, etc. It’s a really unique artifact that hasn’t been documented previously. It’s a pretty fascinating site to be exploring.

The objective of our work there was to provide the archeologists with a baseline map before they started further excavation on the site. We first visited the site in September 2014 and conducted surveys over the main target area – as well as looking at other targets of interest that had been identified using acoustic multi-beam data collected in the previous year.

We returned in June 2015 and mapped a very extensive part of the coastline around the wreck site because some of the diving operations they’d conducted in 2014 identified other areas of interest. We delivered a map of their target excavation sites and they were able to use those maps in situ. They actually had copies of the maps on iPads, and were able to annotate these while diving underwater in depths of 50-60 meters. They were able to log all of their finds and situate themselves on the wreck site relative to the maps that we generated. It provided them with a baseline of what the site looked like before they started with the excavation operations.

They’ve just finished their field research and an exhibit has opened in Basel, Switzerland showcasing some of the finds and describing the site and some of the archeological work that’s been going on there. It’s been a really exciting project to be part of and has attracted a lot of interest. Our part of that was using AUV systems to give them baseline high-resolution 3D maps of the wreck site.

Ron Vanderkley: It kind of sounds like “Jacques Cousteau meets robot”

Stefan Williams: In fact Jacques Cousteau visited the site in 1976 and they did some excavation that was part of their Calypso program. They did a lot of underwater exploration and recovered a few artifacts, but this is really the first time that the site has been systematically mapped.
Ron Vanderkley: What is the future of AUVs? Where are the challenges?

Stefan Williams: AUVs have a big role to play in expanding our understanding particularly in deepwater environments. Our main focus has been in coastal environments linked to ecological and biological studies. I think there is a role to play also in understanding how changes in oceanography (for example temperature, salinity, and current profiles) are affecting the marine habitat. We have a big program as part of Australia’s Integrated Marine Observing System Program, running AUVs to look at how key ecological environments around the country are changing over time. We revisit these areas, and work closely with biologists and ecologists to document changes.

Another big role for these autonomous platforms is in understanding deepwater environments – getting them out into the abyssal plains to find out more about what’s down in the deep parts of the oceans, of which we know very little about.

One thing that really excites me is looking at some of the advancements that were made in long-term autonomous deployment. A lot of these autonomous underwater vehicle systems are really starting to come into the norm. There’s some fantastic work going on around the world in designing and building these systems both on manned surface vessels that are capable of very long range operation, and on underwater vehicles that are capable of thousands of kilometers of deployment.

That really changes the dynamics of how ocean science can be conducted. It reduces the requirement of having people on very expensive ships out at sea for long periods of time if you can send these robotic systems off to work autonomously. We’ve seen the precursors in oceanography, with autonomous float systems being used to help oceanographers get a better understanding of ocean circulation and physical oceanography on a global scale. The next revolution will be autonomous platforms that can do mapping and excavation work over long periods of time in deep water across the world. I think there’s real potential in that area.

Ron Vanderkley: What is Australia’s role in that? Are we able to create our own submersibles? Are we concentrating on the programming in that arena? Where do you see us?
Stefan Williams: We have covered both areas natively in Australia. Some of these vehicles are being purpose-designed for particular tasks. We’re currently designing our next generation of hydrolution survey vehicle for supporting the Marine Robotics program, and there are programs in Queensland and Tasmania looking at requirements for working under ice or in complex reef environments.

There is work here in Australia in building these sorts of systems, though there is not a very large commercial scene yet (a lot of the commercial platforms are being developed in the US and Europe). But there’s certainly a lot of research effort focused on designing vehicles that are fit for purpose, and I think we’re particularly well known for the mapping and the navigation work that we do. We have links with all the major oceanographic groups around the world. I think Australia as a whole is very well regarded in robotics and autonomous systems – both within the marine space, and also in ground and air vehicles – and is having a big impact in this area.

Ron Vanderkley: Finally, I’d like to ask you about employment prospects in Australia. If you were a student going to university looking at the options, where would you point them?

Stefan Williams: We have a lot of students come both through undergraduate and postgraduate teaching and research schemes, and the majority of them do find work both domestically and internationally. There’s certainly a lot of activity currently in the robotics space. Some of the big companies like Google, Apple and Uber have made well documented forays into autonomous driving systems, and they’re certainly employing a lot of PhD students, post docs and research fellows who have been employed in various labs around the world, so there’s a lot of potential there. We have former students and staff embedded in the majority of those companies.

Here in Australia we have a domestic market with a number of small startups that are really doing well in developing robotics systems and sensor payloads. The experience students get – and the type of skills that they acquire through their study of mechatronics and robotics – have application across a wide variety of domains, from traditional engineering work, to software, electronics and mechanical design work … all the way through finance and business studies. There’s a lot of interest in data analytics and machine learning, and a lot of the tools and techniques that we apply in robotics can be transferred across a variety of different domains.

One of my most recent students is now heading up a small Machine Learning group at one of the major banks. The skills and underpinning tools required here in our lab to classify marine imagery have direct application into other areas. When students ask me about what jobs are out there for mechatronics engineers, I tell them that there are always good jobs for good students. One of the keys to that is really being passionate about the things that you’re studying, and I think robotics excites a lot of people.

I teach an experimental robotics course – it’s an elective in the 4th year mechatronics programs here at Sydney University. The things that the students do in the course of the 13-week curriculum are astounding. They put a lot of time, effort, and thought into building these robotic systems. I think they feel very rewarded at the end, and the feedback I’ve gotten has always been very positive.
Ron Vanderkley: Thank you, Stefan, for your time. I hope to speak to you again soon.

Stefan Williams: It’s a real pleasure, Ron. Thanks for helping to promote the activities of robotics here in Australia and overseas.

This audio interview was transcribed and edited for clarity.

ep.193: The Open Academic Robot Kit, with Raymond Sheh

Ron Vanderkley — Fri, 16 Oct 2015 18:22:36 +0000

Transcript below.

In this episode, Ron Vanderkley speaks with Raymond Sheh, who is a Senior Lecturer at the Department of Computing at Curtin University and head of the Intelligent Robots Group.

They discuss his Open Academic Robot Kit initiative, which aims to build an international community of students and researchers, all designing and building robots using a compatible set of electronic and basic mechanical components, plus 3D printed parts. Builders then publish their designs so that anyone can recreate and build on them.

Raymond Sheh
Raymond Sheh is a Senior Lecturer at the Department of Computing, Curtin University. He specialises in the areas of Artificial Intelligence, Robotics and Cyber Security, and has been involved in robotics research since 2003.

Raymond’s current activities include robotics for hazardous environments, surgical robotics and the application of robotic sensing technologies to industrial automation. The former includes a significant education and research outreach component to other universities and high schools through his position on the Executive Committee of the International RoboCup Rescue Robot League competition.

Prior to joining Curtin in 2013, Raymond was with the US National Institute of Standards and Technology, developing standardised test methods for response robots used in hazardous environments. He holds a PhD in Artificial Intelligence from The University of New South Wales, an Honours degree in Electronic and Communications Engineering from Curtin University and a degree in Computer Science from Curtin University.

Links:

Transcript

Ron Vanderkley: Raymond, can I get you to introduce yourself to the podcast listeners?

Raymond Sheh: Hi. My name is Raymond Sheh. I’m a senior lecturer here at the Department of Computing at Curtin University [in Perth, Western Australia]. I have also recently started the Intelligent Robots Group with the aim of doing research into intelligent robots that not only perform their tasks but also can explain what they’re doing and inform us.

Ron Vanderkley: Great! I’ve heard so much about the RoboCup Rescue Confined Space Challenge. Is that something that you can quickly describe to us?

Raymond Sheh: Sure. Prior to coming to coming to Curtin, I worked with the US National Institute of Standards and Technology on developing standard ways for testing response robots — that is, robots being used by bomb squads, search and rescue, hazardous materials and so on. One of the outreach activities of this project was the RoboCup Rescue Robot League competition. This is a robotics competition — an international one — for robots that tackle this very problem: robots going into places where humans don’t want to go in order to perform tasks and discover information.

In fact, the robots that have participated in this competition have already been quite useful out in the field. You might recall the robots that made it to the top of the Fukushima Daiichi Nuclear Power Plant in order to inspect the equipment and see what was still out there, trying to make that area safe. The robots that actually made it to the top came out of the RoboCup Rescue Robot League competition and were then repurposed for the task.

We are tackling tasks that have traditionally required fairly decently equipped universities with a correspondingly large budget. We’re not simplifying them but we’re tackling it at a smaller scale, at a scale that is possible for high schools to replicate. In line with this, we are also developing the Open Academic Robot Kit. This is a family of robot designs where all of the mechanical parts are 3D printable on a low cost printer — that is, a printer costing less than $1000.

All of the other parts are available off the shelf or by mail order anywhere in the world and all the design instructions and source code are available in easily editable form under an open source license. The idea is that anyone, anywhere in the world can download all of the information; they can reproduce the base design for this robot. They can use the competition as an application to inspire their own development and their own research and then they can redistribute this back to the international community so that everyone can build on it.

The aim is to be a catalyst, not just for this application but for bringing interesting, research level robotics, in general, to high schools and the community.

Ron Vanderkley: Is the uptake of participation increasing?

Raymond Sheh: Indeed. We ran our first pilot in the first half of this year with Scotch College and Christchurch Grammar. We also had a school from Canberra participating in collaboration with the University of New South Wales, as well as one from Maryland in the US participating in collaboration with the National Institute of Standards and Technology at the Gaithersburg Campus. We all went to the international competition, which was held in China this year, and we held some demonstrations.
The goal is to have more schools participate in the second half of this year. So that, next year, when the competition is being held in Leipzig in Germany, we would have our first truly international event at this scale, with high schools tackling what was previously university level research.

Ron Vanderkley: This is kind of like a DARPA challenge?

Raymond Sheh: It’s a little bit different to the DARPA challenge. I was actually involved in developing the challenges within the DARPA robotics challenge back when I was working with NIST. The DARPA challenges were deliberately very difficult. There’s a term called “DARPA hard,” because they’re trying to induce such a disruptive improvement in capabilities. You really do require fairly significant resources to devote to this.
The DARPA competition’s aim is to improve capabilities very rapidly and they use the fact that this is a very impressive competition as the motivator for getting people interested outside in the general community. We have similar goals but we tackle it a slightly different way. We present a challenge that is within reach of an interested designer technology teacher and a dozen high school students. They can tackle the problems that we put in front of them. They don’t just have to watch from the sidelines. They’re able to participate, they’re able to actually solve these challenges in ways that are very interesting, in ways that the university researchers may not have thought of. That’s one of the wonderful things about this challenge: everyone can be a researcher, everyone can be an inventor in their area of expertise and then they can share that with the rest of the world.

Ron Vanderkley: Hence an Open Challenge?

Raymond Sheh: Hence the Open Challenge. Of course, the DARPA robotics challenge was also open. DARPA actively encourages the teams to publish and transition their developments outside. We just tackle them at slightly different scales.

Ron Vanderkley: If an enthusiast wanted to find out more about the Open Academic Robotics Kit, is there a website?

Raymond Sheh: Indeed. The website for the open academic robot kit is oarkit.org, Open Academic Robot Kit. We also have information linked from our group, intelligentrobots.org

Ron Vanderkley: What do you think the future is for robotics? Where do you think we’re going?

Raymond Sheh: Well I think that there are a few frontiers that we do need to address in robotics that, arguably, haven’t had the focus that I think they’ve deserved. I’m sure that your listeners are very aware of the various high profile cyber attacks that have been happening around the place. They’re starting to attack what we call “cyber physical systems,” computer systems that have an impact in the real world. The electricity grid is one, but robots, automated buildings and the like are another.

Security in these robotic systems is somewhat lacking, mostly because the intersection between people doing robotics and people doing cyber security hasn’t been as tight as it should be. I think one of the big frontiers that needs to be addressed is: how do we make these systems secure? Here at the Department of Computing we also have a big focus on cyber security and we mesh our activities very closely.

The next area where I have some concerns is that of the engineering of the software that goes into a lot of automated systems. As these systems become ever more complicated and ever more connected and ever more modular, issues with bugs and maintaining difficulties become quite significant. The development of good software engineering practices and infrastructures becomes ever more important, in order to make sure that these systems do the right thing and do what they intend to do. Again, naturally, software engineering is one of the areas that the Department of Computing here has a strong focus on.

When people talk about worrying about the robot apocalypse, I think the problem isn’t going to be that the robots, all of a sudden, decide that humans are a nuisance and do away with us like they seem to do in the movies. I think what is a risk is either some fairly catastrophic bug, a mistake that causes issues, or a Doctor Evil somewhere who breaks into our systems and commandeers them. The scary thing is if they do it well, if we don’t have protection in place, if we don’t have the defenses, we won’t know until it’s too late.

I think those issues are more important than questions about whether artificial intelligence systems are going to become sentient. We should worry more about the people who are already sentient.

Ron Vanderkley: What advice would you give a person interested in robotics? Should they look at a multidisciplinary approach?

Raymond Sheh: A multidisciplinary approach. I don’t necessarily advocate doing what I did. I couldn’t decide what I wanted to do so I did a double degree. In my case it was electronic and communications engineering and computer science, here at Curtin University. Even if you do your degree in, say, computer science or electronic engineering or whatever the case may be, get involved in the robotics club, get involved in open source projects, get involved in other activities that keep you in touch with all of these other areas.
Robotics competitions are actually very good for this. They make you tackle a problem that hasn’t been solved before and give you the freedom to do so in a way that exercises all of these different areas. We don’t care how you build the robot, you just need to solve this problem.

You need good computer science, you need good electronic engineering because you need to be able to make sense of the world and you need to have sufficient power and you need to communicate. You also need good mechanical engineering because your robot has to survive and good art because your robot has to be able to convey this information to a human.

Ron Vanderkley: Where does art fit into the equation?

Raymond Sheh: Well, problem solving is an art. If you think about it, problem solving is an art form. That’s one way of looking at it. When I refer to art in robotics, though, I actually refer more to the consideration of human factors and the fact that this system needs to convey meaning to a human and needs to be able to receive the human’s intent. That requires an understanding of how humans behave both rationally as well as emotionally.

A human system interaction that is well informed from an art perspective will surely be easier for the vast majority of people to interact with than one that has not had that consideration. I think art is very important in that regard. The other aspect of it, the field of art … I think we’re starting to get to the point where the field of art can be augmented by robotics in a way that is assessable economically and technically for many more artists.

One of the things that I hope will happen with the Open Academic Robotic kit isn’t just that it gets picked up by researchers and scientists and engineers, I hope that it gets picked up by artists too. They don’t need to modify any of the electronics or any of the code. Because everything is 3D printed they can make these devices, and make robots more artistic, more aesthetically appealing. They can design them to elicit emotions and to appear as different things.

I think, certainly, that while we may not necessarily want a response robot that looks like a kangaroo or looks like a creature of some sort, having that kind of influence can only be beneficial.

Ron Vanderkley: Thank you on behalf of the podcast for your time.

Raymond Sheh: It was a pleasure. Thank you.

ep.189: Robots and Communication, with Eleanor Sandry

Ron Vanderkley — Fri, 21 Aug 2015 19:04:00 +0000

Transcript included.

In this episode, Ron Vanderkley speaks with Dr. Eleanor Sandry of Curtin University about her new book Robots and Communication. In the interview, we explore human to animal communication and what we can learn from it; human to humanoid robots interaction; and human to non-humanoid robots interactions. Also, we discuss Western and Eastern perceptions of robotics.

Dr. Eleanor Sandry

Eleanor Sandry is a Lecturer in the Department of Internet Studies at Curtin University in Perth, Western Australia. Her research is focused on developing an ethical and pragmatic recognition of, and respect for, otherness and difference in communication. Much of her work explores communication theory and practice by analysing human-robot interactions in science fact, science fiction and creative art.

Links:

Transcript

Ron Vanderkley: Good morning Eleanor, if I can first ask you to introduce yourself to The Podcast listeners.

Eleanor Sandry: Hi there, well I’m Eleanor Sandry. I’m a lecturer in Internet Studies at Curtin University in Western Australia. My research is about communication in general, not just on theinternet and I’m particularly interested in human-robot interactions so that we can understand communication more broadly.

Ron Vanderkley: What got my attention to your subject was an up and coming book that appeared at a conference listing, if you can briefly tell us a little bit about the work on this book?

Eleanor Sandry: Yeah, so the book is called Robots in Communication and it’s been published by Paul Grave in the Pivot Series, which means it’s available online and print on demand. It’s a short format book, 45,000 words. It’s shorter than your average monograph and that’s because I wanted it to be accessible to a range of different people. It’s pitched towards people who are interested in robots, people who are interested in communication – human communication theories – so it’s trying to cater for quite a large audience. What it does is look at human-robot interactions and it tries to unpick them in some detail from a number of different theoretical perspectives.

I’m particularly interested in broadening the size of communication beyond just the transmission of information and ideas of success that surround accurately transmitting it. I’m interested in non-verbal communication; the idea that it isn’t just about what we say, but it’s about us using facial expressions and also movements more generally. I found that robots were just agreat way of looking at a huge range of different types of communication because they are sometimes built to look as human-like as possible, but on other occasions are completely machine-like, and yet what you find is the people who interact with them often develop some kind of relation to them. Even to a Roomba and other iRobot floor cleaners. Just looking at that gives me a whole range of different forms of communication to consider and things I can say about communication in general.

Ron Vanderkley: You speak about this idea of humanoid robots and the non-robots as a point of contact. What actually steers you towards the humanoid? Is it the facial features, the gestures, themovements?

Eleanor Sandry: I think that most people are studying humanoid robots because of particular facial features and expressions, in some cases. But, in others, it’s to do with physical ability to move around in human-like environments. One ofthereasons the Atlas robot – which is about to compete in the darker challenges – is human-like is because it’s expected to go and work in environments that are tailored for humans, but where humans can no longer enter, for example because of some catastrophic disaster.

In fact, I am much more drawn to non-humanoid robots and I’m particularly interested in the way that people will still read pretty much anything that moves as somehow trying to communicate with them. I’m also conscious of the fact that people who are trying to work with robots do better if they do develop this kind of relationship with them. Part of what I have been writing in the book is actually about looking at how those relationships develop and seeing those, say, between humans in ordinance explosive disposal and their robots. Relationshipsbetween people and robots are not necessarily a bad thing or something that we should be worried about; it actually allows those teams to operate better in real world environments.

Although a lot of people say that a sociable robot needs to be able to make gestures, maybe to also have facial expressions; that depends on the structure of the robot. I think there is a lot to be learnt from human-animal interactions. The things that humans and animals are capable of doing together open up possibilities for lots of other forms, for robots to be seen as effective teammates, basically.

Ron Vanderkley: For instance, canine interactions, this kind of trust and understanding. That is a whole different facet and we are still coming to termswith how wecan do that in robotics. Are you picking that apart and trying to figure out how wecan do that?

Eleanor Sandry: Yeah, I’m particularly interested in the idea of the need to have, almost, a trust and respect relationship in order for the team to operate effectively together. In the human-canine relationship, there would always be an overarching power relation. The human is almost always in control of the overallsituation and is working with the dog to achieve a particular purpose, for instance tracking a scent trail. The same is true with humans and robots and, at the moment, say with ordinance disposal robots, broadly, the human is in control almost 100 per cent of the time: they’re still mainly radio controlled or controlledvia wire connection. There are some things that they can perform autonomously like righting themselves if they fall over and maybe also returning to base.

What I’ve been trying to unpick, I suppose, is the idea that if a robot is going to become more autonomous, it will be out of contact with its human controller. There needs to be a more complex relation between the human andthe machine in order for that team to operate well. Thehuman effectively has to relinquish control to the machine at the point of which it moves out of radio contact, they have to trust in the robot to do its job. All of these ideas – even the terminology of me saying to do its job – isproblematic, historically, in understanding human-animal relations. Because many people think that a dog doesn’tknow what its job is, or isn’t going to take responsibility, and yet people who train dogs will tell you that this is precisely the way that you need to think about the dog in order tounderstand what is happening in the team, because you don’t have the nasal senses that the dog does.

It’s actually important that you trust a dog to do its job, so when it appears to be moving off a trail, in fact its followinga path that you have no idea exists and it will bring you to the final point; will complete the task effectively. I think thereare possibilities of the same thing between human robot teams, where the robot has senses that a human doesn’t. You have to learn to trust in those senses and I think it’s an interesting way to think about these relations. I think it has practical possibilities in the robotics community, but I also think it tells me all sorts of interesting things about how humans and other things interact.

Ron Vanderkley: The idea that we could try to use speech recognition between robot andobject, that being the only means of communication in the animal or non-animal world, is not entirely true. It’s looking at gestures, our voice.Is this also something that we may not have looked at in the robotic world?

Eleanor Sandry: Well I think there is a tendency. It’s not so much that it hasn’t been looked at but that the final goal is seen as being able to create a robot that it’s possible to activate entirely with just your voice and that will understand human expressions and gestures perfectly. I think that what’s being missed is that the robot could understand things differently from us. It’s not actually necessary to embed the robot in the world ina human-like way. It can be embedded in its own way and understand us in particular ways, even if that’s using sensors on hands that help direct the robot. Actually looking at those as real possibilities, not a problem that needs to be overcome, has potential forincreasing the flexibility of the way robots work.

The other reason I’m particularly interested in that is because I hope to start writing more about critical disability studies. That has really opened my eyes to the fact that technologies that are currently being directed towards very human-like interfaces may actually make it harder for people with disabilities to interact with a robot that might otherwise be very helpful to them in living their everyday lives independently. Really looking at a whole range of different interfaces for machines like this is important; not just thinkingof trying to perfect this notion of the human-like interface. I just don’t want them to cut down possibilities I suppose.

When I talk to people sometimes about robots, I get two reactions: normally, I get the people who are really firmly interested in human-like forms and want to make sociable robots as human-like as possible, and then I sometimes get other people that say “we never make robotshuman-like.” I get this divergent reaction and, really, I’m somewhere in the middle, trying to say “well, can’t we think about this differently? Can’t we think about this froma communication perspective, where you start seeing all the different ways that people and things and animals can communicate?” Because you’re quite right, of course, dogs don’t just respond to voice, they respond to tone of voice. They also understand things in ways that we can’t really possibly fully appreciate because wejust don’t have the senses that a dog does.

Ron Vanderkley: The other aspect I was considering is that we are talking about robots actually understanding us. I spoke to a guy in the UK about, I think, a robot called Hector, where it’s entertainment. It’s a robot telling the general public what is going on, with hand gestures, eye movements, et cetera. It’s a teaching tool, almost. I found that very interesting. Is that also something that you’re looking at?

Eleanor Sandry: I suppose it’s not my primary area of interest but that doesn’t mean I don’t think it’s important. I think that sometimes when I talk it’s easy for me to get obsessed with all the non-humanoid robots and it makes it sound a bit like I’m actually saying that we shouldn’t have humanoid robots and I don’t think that’s correct at all. I’m sure there are going to be situations where a humanoid robot works much better and that is probably a very good example. If you have a robot that you want to actually present to people and you want, effectively, to present it in a human-like way, then that is what you’re aiming for. You’re aiming for something that inspires the audience to pay attentionto it because of its abilities – that may sometimes be very important.

I think that thinking that that’s the only way to build a robot – that can be captivating, or attract attention, or draw people into communication – is probably incorrect and there are other examples. Certainly robots in art, for example, where a totally non-humanoid machine completely captivates its audience through its movements and the things it does, the sound it makes. It’s a very different idea; it’s not to try to transmit information directly but it’s still building a kind of relation between the human andthe machine that may lead us to question the ways that things exist in the world, which I think is usually quite a good thing. Also when things are framed by particular tasks, so anything where people and the robot aretrying to complete a task together kind of frames the communication that takes place.

I think it offers many more possibilities for non-humanoid communication in that particularsituation because you know what you’re aiming to do. And yourcapability of reading non-verbal communication is probably much greater in that situation. With the example of the robot up on stage, often a humanoid robot is going to win the game there.

Ron Vanderkley: When you look at the history of, say, robot interaction targeted at children or things like the first robot dog, where children actually relate to that almost as a real dog, and treat it as a real animal, do you think that’s the precursor to a closer relationship with non biological entities?

Eleanor Sandry: Well, I think its one direction in which to go. I’m slightly concerned by the idea of trying to create robots that effectively replace a relationship that someone would have with something that already exists in the world. It’s one of the reasons that I’m not so interested in humanoid developments, and also in the very animal-like robot developments, because it may notbe a good thing to replace the relationships that people already have with living things with a robot equivalent. Children are actually a fascinating example of humans who are freed from the need to justify all the decisions they make about how they’re going to interact with something and so, yes, theyreact to things, like a robotic dog,strongly. Their understanding, if something goes wrong with it, that it needs to rest or sleep, shows just how strong their understanding of the robot is.

I think that’s one of the reasons why I’m interested in robots that appear more overtly machine-like, because they remind you that you’re in quite acomplex new relationship with something that’s different. It’s different froman animal, it’s different from a human; it is its own thing.For example, I looked at interactions between humans and Guy Hoffman’s robot, which is a robotic desk lamp, a lighting assistant, and they did experiments together. People’sinteractions during the interaction were very much as if it was somewhat alive, not necessarily like an animal or anything else, it was an alive desk lamp. But you could still switch it off; it was still a machine as well.

Humans are capable of holding both of those ideas. I think children may be less worried about that, so as a specific example they do show just how far those relations can go. I suppose I’m interested in how you can signify that those relationships are actually different, that it is still a machine. However, I don’t really think that robots should replace human-animal relations. I just read something recently about a different researcher in Australia who has written about how he thinks, in the future, humans will have robotic animals and that immediately made me think of Doandroids dream of electric sheep? The original book has a very strong section in the front that is about the ownership of robotic animals being all that was possible for most people. The richest people had a real animal. It seemed to be that future idea and I hope and pray that we are not going to go that way.

Ron Vanderkley: No. There’s a concern throughout the entireinternetaboutthe future of autonomous robots, AI being this Skynet-kind of world, by those who I think are quite smart people. It kind of sends the wrong signals. I just wanted to see your take on things.

Eleanor Sandry: Well, my take on it is artificial intelligence has a long way to go before it ever getsto that kind of level. I also think that the way that it’s being developed at the moment, it’s not necessarily ever going to be the same as a human-like intelligence. An artificial intelligence is going to be different almost certainly. Yes, I get frustrated at the hype around the technology, and you’re right, it’s very respected voices that are raising these kinds of issues,and in the near future as well. I simply don’t think it’s going to happen in the near future and I don’t actually think it’s going to happen the way that some people are suggesting it might either.

I would definitely want to draw back from that and point out just how far away artificial intelligence is from being at that level. I mean Google Drive. These cars don’t drive everywhere; they drive in one particular place, where it’s been mapped sufficiently for them to be able to operate. I don’t think that people are necessarily aware of that when it’s reported in the mainstream media.

Ron Vanderkley: Not at all, and humanoid robots actually cost an awful lot of money to build …

Eleanor Sandry: Huge amount.

Ron Vanderkley: I don’t even know if Sony is actually able to clear a real amount of money.

Eleanor Sandry: I doubt it very much. I mean, that robot has been in a series of developments, way over a decade now, and still has issues being placed into a normal physical environment and operating successfully. It has to be ina controlled environment. I think that the general public probably isn’t aware of that. The difference between seeinga robot acting in a laboratory, in a controlled setting, and actually being able to interact in the normal physical world that we live in is something completely different. They are streets apart.

I saw something about the darker challenge basically explaining that they would be very impressed if one of the Atlas robots, or any of the other bipedal robots, actually manages to make it through the course without falling down. I think that that’s a realistic understanding of what’s possible at the moment. The idea that robots with four legs might actually operate more easily in those types of environments is fair and should be perceived as well. It’s a very interesting question.

Ron Vanderkley: Absolutely, and just in a previous podcast I was talking to some guys about soft robotics. I mean that’s completely going …

Eleanor Sandry: Totally, a very fascinating area.

Ron Vanderkley: Yeah, completely, a different direction and its early days.

Eleanor Sandry: Yeah, very early days, but there is so much potential for soft robotics but also mixtures of soft and hard. I saw ages ago that, rather than using a hand-like gripper to pick things up, someone had developedsomething that could actually… it was almost like something filled with coffee.

Ron Vanderkley: Like a bean bag.

Eleanor Sandry: Like a bean bag yes, that could grip. Those sorts of innovative ideas that are thinking of totally different ways of solving the task, rather than trying to make something that has a human-like hand, seem to me to offer tremendous potential and so I’m fascinated by the idea of what a robot might end up looking like in the future. I’m kind of encouraged that maybe robots in the future won’t look like us and actually that would be a really good thing.

Ron Vanderkley: That’s another human concern. Someone once put it that if we try and design something that looks so close to a humanand almost mimics the face, there’s total distrust. Whereas if it looks like a robot there is more trust.

Eleanor Sandry: I think there definitely is potential. I mean whenever I write about robots I always end up having to utilize the uncanny valleyexample. I’m kind of quite frustrated with it now: that idea that, as a robot becomes more and more human-like,eventually it’s actually read as zombie-like and there is a strong drop off in the level of trust in the robot. Some people suggest that you could climb the valley on the other side and eventually end up with aperfect human-like machine. Then, as has been written about in many different science fiction books, you have the question of what is the difference between the robot and the human and …

Ron Vanderkley: Silicon versus Calvin.

Eleanor Sandry: I think, really, we’re nowhere near achieving that at the moment. Even the very, very human-like faces you can see pretty easily from their expressions and from other things about them. I don’t know why we want to aim for that. I think that we have …

Ron Vanderkley: Perfection.

Eleanor Sandry: Yes, but it seems to be a perfection that’s driven by a100 percent understanding of humans being the pinnacle of evolution and, therefore, that should be the pinnacle of robotic development. That just seems to me an unfortunate decision to make. There are many more options for robotics than trying to achieve human-like status. I think that there are many other ways that machines could be far more interesting, far more useful but also maybe even develop new relations with people that don’t have to be like something that already exists.

Ron Vanderkley: One of the ideas in modern day sci-fi, the cyborg…

Eleanor Sandry: Yes, the idea of augmenting human ability, of fixing people, which is a very hot topic in disability studies because,in general, there aremany people with disabilities who do not wish to be fixed.Again, it’s driven by this idea of the human being the current pinnacle of evolution. Let’s see howfar we can take it. I hesitate to mention this but, as Dona Harrower said many, many years ago, in 1985 actually, we are already cyborg. I’m effectively a cyborg: I wear contact lenses; that’s the only way I can exist in this world. I would not be alive without those developments – of glasses or contact lenses. But where people begin to decide to amputate limbs, in order to have blades, because that means they can ran faster, that’s another thing science fiction is written about. There are all sorts of – not just ethical – questions, but bigger questions around that.

Ron Vanderkley: You brought up sci-fi as a sounding board. Do you use that as a sounding board for some of your ideas or your thoughts?

Eleanor Sandry: Often I do, yes: in fact, in robots and communication. Science fiction to me is a really great place to look because it’s sometimes overlooked as being a kind of subgenre. For some people, for many years, it was not real literature and not something that was studied much in scholarly circles. But it provides a real sounding board and some of the best science fiction writers write about embedding technology in societiesin ways that allow you to really think about some of these questions ina more developed way,because someone has written a story about it. It seems to provide a reallygreat way of thinking through some of the questions that are very difficult to picture in your head, whereas science fiction provides you with those kinds of thought experiments that can be so important.

In my thesis – what I was doing for my PhD- I did write more about science fiction: the idea of culture and human-machine interactions in culture were to me very interesting because they push way into the far future I guess.

Ron Vanderkley: Coming back on the idea that technology is going to take over. Is that something that you think that’s going to be more and more of a problem, as sci-fi seems to dwell on the bad things and not the positives, or is that just good for entertainment?

Eleanor Sandry: Well, it’s definitely good for entertainment. It’s also a very Western perspective on robots andtechnology. Whereas in Japanese Manga, the whole idea of technology, and robots in particular, is far more positive. Also, it’s a great pity that we lost [Iain] Banks because that was someone writing about a utopian feature, problematic in lots of ways, but at least he offered some kind of positive vision of the way things could work. You’re right, in general, there will still often… they’re not exactly disaster movies are they? But they’re kind of …

Ron Vanderkley: The good guy always seems to win in the end.

Eleanor Sandry: In the end but there is usually a potential for total disaster in the middle, isn’t there?

Ron Vanderkley: Okay, so in closing, what do you think is going to happen in the field of robotics, well, in Australia?

Eleanor Sandry: Australia is interesting and I’m still trying to get to grips with what’s happening in different parts of Australia. I think that the level of funding here may be very different from other countries. There are interesting things going on and I really need to spend more time getting to grips with them because I keep on reading things about what people are doing in Melbourne, starting with robotic arms, the disabled, people with disabilities, and I’m interested that there is stuff going on in Sydney. I’m seeing people from the Center for Social Robotics there, and so I’m particularly interested in what they’re doing and the cost as well.

I think robotics will continue to develop here. There is probably quite astrong drive towards thinking of robots in terms of self-driving machines, mind technology. That’s always going to be a big thing in this country. There are definitely innovative thinkers in Australia who produce interesting solutions to different problems and I would like to carry on finding out what they’re doing and writing about it.

Ron Vanderkley: Okay and to cap that off, if I can firstly thank you for giving the time to chat. On behalf ofThe Podcast, thank you very much.

Eleanor Sandry: Well, thank you for having me. It’s been great to talk.

All audio interviews are transcribed and edited for clarity with great care, however, we cannot assume responsibility for their accuracy.

Rehabilitation and Environmental Monitoring, with Lei Cui

Ron Vanderkley — Fri, 26 Jun 2015 11:03:12 +0000

Transcript included.

In this episode, Ron Vanderkley speaks with Dr. Lei Cui from Curtin University about his team’s work on 3D printable hand orthosis for rehabilitation, a task-oriented 4-DOF robotic device for upper-limb rehabilitation and a 3-DOF platform providing multi-directional perturbations for research into balance rehabilitation. They also discuss a high-speed untethered robotic fish for river monitoring and an amphibious robot for monitoring the Swan-Canning River System.

A 3D Printable Parametric Hand Exoskeleton for Finger Rehabilitation

ComBot: a Compact Robot for Upper-Limb Rehabilitation

A 3-DOF Robotic Platform for Research into Multi-Directional Stance Perturbations

Curtin Robotic Fish

AmBot: A Bio-Inspired Amphibious Robot for Monitoring the Swan-Canning Estuary System

Lei Cui

Dr. Lei Cui completed his PhD degree in Mechanical Engineering in July 2010 from the Centre for Robotics Research, King’s College London (UK). He continued his employment as Postdoctoral Research Associate after graduation. In February 2011 he moved to the Robotics Institute, Carnegie Mellon University(US) and worked as Postdoctoral Fellow until July 2012. He joined the Department of Mechanical Engineering, Curtin University (AU) and was appointed Lecturer in Mechatronics, which he currently holds.

Links:

Transcript

Ron Vanderkley: Dr. Lei, we can start by asking you to introduce yourself to the listeners.

Dr. Lei Cui: Sure. My name is Lei Cui, I’m a lecturer in Mechatronics. I’m with the department of Mechanical Engineering at Curtin University. I joined Curtin about two-and-a- half years ago. Before that I got my PhD from King’s College, London, and I did my second Post-Doc at Carnegie Mellon University.

Ron Vanderkley: Let’s talk about the research that you’re doing at Curtin University. I know there are two facets to it. We can start with the first one: robot rehabilitation.

Dr. Lei Cui: I worked on robot rehabilitation when I was doing my PhD. After I joined Curtin, I decided it was a good time to pick this research up again, because the Mechanical Engineering Department is now shifting its research focus to rehabilitation and medical engineering. In the last two and a half years I’ve developed a 3D printable hand exoskeleton for rehabilitation, in collaboration with a colleague. We patented a finger design just three months ago. The unique feature of this hand exoskeleton is that each finger has three active joints, and, so, three active outputs, but with only one actuator – there’s one input. It’s the first design of its kind in the world.

It’s a joint linkage, I don’t know if you’re familiar with the concept of linkage? We applied the same linkage to all five fingers, so each can be operated by one linear actuator, but has three output joints corresponding to the three joints in each finger. Another feature is the whole assembly can be printed out using a 3D printer. No assembly work needs to be done, it’s for mass customization.

Ron Vanderkley: That’s a general 3D printer, printing in any kind of polymer plastic?

Dr. Lei Cui: No, we need a high precision printer. We need high resolution.

Ron Vanderkley: It’s not a hobbyist or …

Dr. Lei Cui: No, it’s not, any low level 3D printer cannot do that.

Ron Vanderkley: You told us that it’s at the point of being patented or patented already, are you collaborating with any other organization?

Dr. Lei Cui: Yes, we’re now working with a company in the US, from MIT. They’re also developing this hand exoskeleton, but our design is much better than theirs so we’ve been talking about how they could buy our design, or we could just print some collaboration frames so we can work with them.

Ron Vanderkley: You also spoke about the second project that you’re currently working on?

Dr. Lei Cui: Another project is an upper limb, for arm rehabilitation. On this project I have been working with Kerry Allison, a professor at Curtin from the Health Science School. Mainly, we’ve concentrated on task-oriented training. When patients do training they don’t focus on each individual joint in terms of rehabilitation. Usually, they focus on the task. For example, they want to move the hand from point A to point B: they normally look at the target first then move the hand from A to B. They don’t focus on each individual joint.

So that’s the idea behind our design. The ideal training would be for the user, or the patient, to look at the target, move to it and it would have false feedback to test the patient’s performance.

Ron Vanderkley: So it’s a variable device?

Dr. Lei Cui: No, it’s not variable, it’s a hand infector type. Not an exoskeleton type.

Ron Vanderkley: The primary use for that is to teach stroke victims?

Dr. Lei Cui: Yes, mainly for stroke patients.

Ron Vanderkley: What stage are you at with that process?

Dr. Lei Cui: We have developed a prototype and, currently, I’ve got one PhD student working on this project, but he’s only at the first stage. Now, we are combining force torque sensing and EMG sensing to do a patient intention estimation, it’s the first step. Previous work has used a predefined trajectory, meaning that the subjects first perform the task and they record the predefined trajectory; so they’ve try to force or assist the patient to follow the same trajectory. Our idea was to do a patient intention estimation first and combine EMG and force sensing. So, we don’t rely on the predefined trajectory, we want to generate it on the fly. This is the first step.

The second step would be to apply adaptive impedance control to assist training, as needed.

Ron Vanderkley: Thirdly, you are working on a three directional platform, is that correct?

Dr. Lei Cui: Yes, a three dimensional platform. That’s also a collaboration between Professor Gary Alison and myself. Basically, it’s his idea; he’s been working on this project for many years. What he wanted was a platform that could be multidirectional for the patients. This platform destabilizes the user. So, the user will be standing on top of the platform, it will move and the user has to move his or her foot to regain stability. That’s the part we need to control well.

Ron Vanderkley: What’s the end goal of it? Why are you creating it?

Dr. Lei Cui: Professor Alison’s research focuses mainly on how to prevent old people from falling. To study human posture and balance, you need to provide a well-controlled multidirectional platform for patients. That’s why it’s pretty large: 600mm in diameter. A person with an average weight of 100kg can stand on top of it comfortably. Also, they can shift a foot and regain their balance; we’ve provided enough space for that, and also enough speed and acceleration.

Ron Vanderkley: The other area of expertise you’re looking at is Biomimicry, could you tell us where you are going with that?

Dr. Lei Cui: Biomimicry is a fascinating research field; it’s also quite a hot topic. If we take the normal approach, using mathematical modeling, it’s very hard to realize some functions, that’s why we take inspiration from nature. Because nature – fish and other animals – has evolved over millions and millions of years. I’m focusing on applying Biomimicry design to monitoring.

The first project I worked on was called Amphibious Robot. We took inspiration from amphibious centipedes because they have multiple feet. These feet provide buoyancy and actuation, our design was decided with this in mind. It’s a track driven vehicle, but at each point of the two tracks, we’ve fixed ballpoint blocks so that, when the robot is in water, they provide buoyancy and actuation. When it’s on land, it can also effectively drive the robot.

Ron Vanderkley: At what stage are you at with that?

Dr. Lei Cui: We have finished the first stage, that’s the mechanical design and the empirical study; the paper’s been accepted by the ASME Journal of Mechanical Design. Now, some of my students are designing autonomous navigation systems, using an android form that also provides sensors, an accelerometer, gyros and a video camera.

Another project is the Robotic Fish.

Ron Vanderkley: Yes, we’ve heard a lot about the fish.

Dr. Lei Cui: In fact, this robotic fish is quite fast. It’s the fastest robotic fish in the world in terms of body length, per second and we’ve achieved three body lengths per second. A review paper in 2012 held the previous record: 1.6 body lengths per second.

Ron Vanderkley: So, it’s a first here in Western Australia?

Dr. Lei Cui: Yes, definitely, we just haven’t published anything yet. We want to wait until we collect enough data and it’s accepted by a good journal.

Ron Vanderkley: The propulsion of the fish…Biomimicry, is that using actuators?

Dr. Lei Cui: Yes. In fact, those kind of fish are in [the local] Swan Canning River, they’re called Bream. It’s quite a popular fish in this river. We effectively morphed this fish. If you put our robotic fish in the river, it would be very difficult to discern which is the real fish. The appearance is the same and the size.

Ron Vanderkley: Is there a plan to monitor the schools of fish?

Dr. Lei Cui: Yes. Our idea is that our fish could mix with biological fish seamlessly, because they could be made to look real. We hope, in the future, to use it to monitor the fish population in the Swan Canning River, for example.

Ron Vanderkley: Would you be able to mount sensors on them to take salinity tests, for example?

Dr. Lei Cui: Yes. Now we’re studying the actuation of the fish; we hope to further improve the speed, that’s the first task. The second task is to develop electrical field sensing technology for the fish, because they are un-tethered, small in size and battery powered. Currently, the batteries last for about 45 minutes but we need to make the fish autonomous. We’ve taken inspiration from some fish using electrical fields to sense their environments. I’ve got one PhD student and several undergrad students developing this form of bio-sensing, using electrical field sensing.

[0:17:53]

Ron Vanderkley: That’s very fascinating and impressive. I haven’t read anything on that for quite a while.

Dr. Lei Cui: Thank you.

Ron Vanderkley: We’d be very interested to find out how that’s developing, once you get going on it.

Now, the question that we ask everybody is this: Where do you think robotics will be, say, in 10 or 20 years from now? What kind of direction do you think we are heading in?

Dr. Lei Cui: I think robotics covers a lot of areas: artificial intelligence, mechanical design, sensing technologies; many of them are a combination of these. In my opinion, currently, sensing technology is holding back development in robotics. I would assume that, in the next 10 or 20 years, more people will be focusing on sensing. In terms of artificial intelligence, IBM has built machines to beat world chess champions. Artificial intelligence to do a single task is, I think, quite mature, but to multitask? Developing robotic animals that behave like the real thing still has a long way to go, it needs 50 -100 years, I’d guess.

Ron Vanderkley: Obviously, we’ve seen lots of examples, but predefined scripts give the audience the idea that some robots are more intelligent than they are?

Dr. Lei Cui: I’ll give just one example. While I was working at CMU, on the DARPA sponsored ARM [Autonomous Robotic Manipulation] project, one of my tasks was on the robot arm to hold a key, insert it into a lock and then unlock the door. If you watch the video, it seems quite intelligent. It appears as if it can feel what it’s doing, but my approach was to use force torque sensing on the wrist of the robot and then to insert the key into the hole. It looks intelligent but, in fact, as you’ve said, it’s defined by algorithms.

Ron Vanderkley: To me, that always confuses people and detracts from the research, that should be the focus. It presents the wrong impression. Similarly, movies like Terminator, also paint the wrong picture.

Dr. Lei Cui: Yes, definitely. I think it would take 200 years, at least, for Terminator to be a threat.

Ron Vanderkley: So, I’ll wrap it up by saying thank you very much for taking the time to do the podcast, and we hope to chat with you again in the future.

Dr. Lei Cui: Sure, thank you.

All audio interviews are transcribed and edited for clarity with great care, however, we cannot assume responsibility for their accuracy.

ep.180: Soft Robotics Toolkit, with Donal Holland

Ron Vanderkley — Fri, 17 Apr 2015 15:34:00 +0000

image: Soft Robotics Toolkit

In this podcast, Ron Vanderkley speaks to Donal Holland of Harvard University about his team’s work on the Soft Robotics Toolkit.

Soft Robotics is a class of elastically soft, versatile, and biologically inspired machines represents an exciting and highly interdisciplinary paradigm in engineering that could revolutionize the role of robotics in healthcare, field exploration, and cooperative human assistance.

The Soft Robotics Toolkit is a collection of shared resources to support the design, fabrication, modelling, characterization, and control of soft robotic devices. The toolkit was developed as part of educational research being undertaken in the Harvard Biodesign Lab. The ultimate aim of the toolkit is to advance the field of soft robotics by allowing designers and researchers to build upon each other’s work. The web site contains the open source fluidic control board, detailed design for wide range soft robotic components (including actuators and sensors).

https://www.youtube.com/watch?v=67qhphkkibo

The growing popularity of site is now bringing in hobbyist and makers alike. The Soft Robotics Toolkit team has announced two competitions intended to reward students, researchers, makers, and designers of all levels for their contributions to the field of soft robotics.

Donal Holland

Donal Holland is a visiting Lecturer in Engineering Sciences at Harvard School of Engineering and Applied Sciences Demographic info Ireland | Mechanical or Industrial Engineering. He was a passed PhD Student at Trinity College Dublin, Visiting Fellow at Harvard School of Engineering and Applied Sciences, Research Assistant at Treocht Ltd.

Links:

Transcript

Ron Vanderkley: Welcome to the podcast. Can I get you to introduce yourself?

Donal Holland: My name is Donal Holland and I’m a visiting lecturer at Harvard School of Engineering and Applied Sciences. I’m a mechanical engineer, and I’m one of the developers of the Soft Robotics Toolkit.

Ron Vanderkley: Can you explain what soft robotics is?

Donal Holland: When we say soft robotics, we mean robotic systems that are composed partially or entirely from soft, low modulus materials. In plain language this means materials that are very stretchy or those that are flexible or bendy. This is still quite a broad definition, and in our work we mostly focus on actuators and sensors that are composed from fabrics or elastomers, like silicone. That’s what soft robotics means for us.

Ron Vanderkley: Can you explain what the Soft Robotics Toolkit is?

Donal Holland: The Soft Robotics Toolkit is an online connection of resources that are related to soft robotics research and development. We intended it to be an intellectual toolkit rather than a physical collection of parts or tools. The toolkit is a website that describes the design, fabrication, modeling and testing of soft robotic components. We currently have documentation for three categories of components on the website, covering actuators, sensors and control hardware. And for each component, the documentation discusses the design of the component, for example, by giving an overview of how it works, discussing the effect of different parameters on its performance and, where appropriate, we provide tutorials on modeling the component in CAD software. It then talks about fabrication in step-by-step tutorials for how to build this yourself, including building materials and where you can order the materials required.

Depending on the component, we also talk about the kind of research that’s been done on it. For example, how we model this mathematically or numerically, or how we carry out testing to understand its performance. It’s said to be very self-contained and to give an overview of how it’s designed, how it’s built, and how it’s used.

We don’t really talk about how we then combine them into full systems; we leave that open ended to the user. We want to see where people will go themselves. But we do provide some case studies of what other people have done using this component.

Ron Vanderkley: So basically the toolkit has helped students kick start in the field of soft robotics, eliminating the need to join all the dots together?

Donal Holland: Exactly. We want to help people avoid reinventing the wheel and build upon each other’s work so that we can move soft robotics forward at a faster pace.

Ron Vanderkley: As you pointed out, it’s being developed as an academic tool. Has the up tide meant that now there are a whole bunch of other types of users that are very interested in developing their interest in soft robotics?

Donal Holland: That’s absolutely the case. Originally our goal for the toolkit was for it to be an educational resource for use in our classrooms. We teach some design courses and we want to develop resources that could be useful for our students. So we carried out research to understand what our students’ learning needs were, and then we developed an initial version of the documentation and tested it with a range of undergraduate students just to ensure that it was clear enough that they could understand it.

Next we used the results from these tests to improve the documentation, and then we used it in our own design courses for two years. We usee the feedback from these tests and from our own students to further improve the website, and then in September 2014 we launched it to the public.

While our initial goal was to develop a classroom resource, we also realized along the way that the website could serve a broader educational purpose. For example, soft robotics is a very multi-disciplinary field. We have chemists, mechanical engineers, biologists, electronic and computer engineers all coming together, bringing their own expertise together to develop new systems. The toolkit is a way for these different disciplines to share their knowledge and educate each other so that the field can move along at a faster pace. We’ve also had a lot of interest from high school and middle school teachers and students who would like to use it in their classrooms. There’s a big and active hobbyist community around robotics and Makers in general, and we’ve had a lot of interest from people in those communities as well.

Since launching it’s become a lot broader than we originally intended it, and we’re very excited to see where it goes next.

Ron Vanderkley: As you pointed out, it’s essentially a list of components etc. Has anyone suggested turning it into something someone can actually purchase? Or is it left to the individuals for source material?

Donal Holland: At the moment it’s very much left to the individual, but we have had a lot of people suggest or ask if they can buy it. It’s not something we’ve done yet but we’d definitely be open to bringing the components together to make it easier for users in the future. I think one of the big challenges, especially when you’re doing this internationally, is that what might be easy to obtain in one country is almost impossible in another – that’s something that we’re really running up against when we’re working with people around the world.

Ron Vanderkley: I was wondering about how to come up with some of the components in this part of the world …

Donal Holland: I had that experience myself when I lived in Europe, and there’s a lot of really good open source, open design projects out … there but a lot of the time they are based in the US and trying to find the same components in Europe is very difficult. I understand the frustration that people have when they’re trying to do that, so as an interim step we’re very keen to work with people in other countries to find local suppliers that could do equivalent parts. For example, we’re working with some educators in Brazil who are sourcing parts in their own country that we will then put up on the website. We are very open if anyone in any part of the world is interested in contributing lists of local suppliers – we’d love to host that on the website to make it easier for people to be involved.

Ron Vanderkley: What were the challenges in creating the kit?

Donal Holland: One of the big difficulties with sharing this type of information – where you’re trying to take expert’s knowledge and share it with other people – is the problem of tacit knowledge. People who are very familiar with a particular field or technology or fabrication method typically have a lot of knowledge that it’s very difficult to make explicit. They often don’t know that they know themselves, if that makes sense. And so it can be difficult for experts to document their knowledge in a way that’s sufficiently clear for non-experts to understand.

Put in another way, an expert has forgotten what it’s like to not be an expert, so it’s difficult for them to educate other people sometimes. Our way of getting around this has been to observe interactions between students and experts and to record a knowledge that is shared. The great thing about these interactions is that the experts have to try to explain their knowledge to a beginner and the beginner is able to ask questions about anything they don’t understand.

That’s really useful for our purposes, and it’s how we overcame that difficulty when we were creating the tool kit.

The next big difficulty that we have is for this to really be useful and sustainable … We need to encourage experts from all around the world to share their knowledge. We want this to be much bigger than ourselves and we see this as a community resource – not our own website. We’re trying to figure out a way to incentivize people to be involved to contribute their knowledge and to stay involved.

Ron Vanderkley: Are there example projects that are being created that students can actually take a look at, or get inspired from?

Donal Holland: The toolkit focuses very much on the basic components, which can then be combined to create new devices and systems. By focusing on the fundamental building blocks and keeping it open ended, we hope that more people can take it places that we never would have thought of. To encourage this we’re currently holding a soft robotics design competition, where we’re asking website users, including students and hobbyists, to create something using materials on the website and then to document their designs. Then in July we’ll announce the best designs, share them on the website, and the designers will receive cash prizes. The documented designs will act as examples of what people can do with the toolkits that students can learn from.

Ron Vanderkley: What is the future of the Soft Robotics Toolkit? Has someone set a timeline? Is it just going to run almost by itself?

Donal Holland: Ideally, we would like it to get to a stage where it could run by itself. I think right now we’re still very much figuring out what the future of the toolkit is. It’s only really been a few months since we launched to the public and we’re still basically collecting feedback from people to figure out what would be most useful for the soft robotics community, what would be most useful for the maker community, and of course what would be most useful for students. We’ll use that feedback to try and figure out where we go next. But I do think a big part of it is trying to make it more of a community resource and people who are outside the original development team taking over responsibility for parts of the website and making it their own.

Ron Vanderkley: What is the future soft robotics? Where will it go? Specifically, I noticed that the whole kit is dependent on quite a large environment. Has anyone looked at miniaturizing it and tethering it from the rig?

Donal Holland: That’s a great question.

In terms of the future of soft robotics, I think there’s a lot of excitement about the field … so I’m really interested to see what the near future is going to hold. Our main interest is in medical applications – most of the research we do is related to the medical applications of soft robotics – and there are already some very nice soft robotic medical devices on the market. That’s one very interesting area that’s definitely going to continue over the next few years. Another area that’s receiving a lot of attention is industrial applications, so in particular end effectors for grasping will definitely continue to grow. In both these situations I think it’s okay to be tethered. For example, if you’re on an industrial robot with a large robotic arm, I think it’s okay to have a fairly large pump that you can be tethered to. You don’t need to be too mobile.

In terms of research applications, a very exciting area right now is biomimetic robotics. There’s a lot of really cool work being done on robots that imitates biological forms – especially animals. One of the reasons this is so interesting is that, on the one hand, roboticists are learning from biologists and are making better robotic systems as a result. And on the other hand, I think robotics research can potentially lead to a better understanding of biological systems. But this is an area where making things untethered becomes very important … for example people building autonomous swimming robots. There’s some pretty good work being done in Daniela Rus’ group at MIT in this area; they have completely un-tethered fish robots that can swim around without being attached to a big rig. This is the area where untethered systems are really going to come from, and there’s going to be some really interesting research done in the next few years.

Ron Vanderkley: I noticed that they had an example of a combustion system that uses gases to generate thrust. Is that just another up and coming example of where soft robotics can go?

Donal Holland: This is an area I haven’t been involved with really. The documentation for that project came from ETH, which is great – we are delighted to have contributions coming from outside, and that was our first contribution. There is some work being done at Harvard in that area. I think it’s really interesting because in typical pneumatic systems, the actuation time is quite slow and in our applications, which are often medical, we’re not really bothered because speed isn’t that important – what’s more important is that you’re not doing any damage.

But for biomimetic forms, if you want to have a robot that can jump for example, then you want to be able to do much faster actuation … and that’s where combustion becomes really, really interesting. Again, I think that’s where the biomimetic robotics is really pushing the boundaries of soft robotics actuations. It’s definitely something to keep an eye on and see what’s going to come out that area in the next few years.

Ron Vanderkley: Should we be looking at the Harvard lab for their next steps, or are there other organizations also moving as rapidly as Harvard?

Donal Holland: I think there are a lot of groups doing really good stuff around the US, almost too many to list. There’s a lot of really great work at MIT, Carnegie Mellon, at Cornell, a lot of interesting wearable stuff going on at Stanford, in ASU. Around Europe there’s of course the RoboSoft CA that are coordinating a lot of the soft robotics research in Europe and further afield. So it’s hard to try and list off the groups here because there’s just so many it’s really grown in the last few years.

Ten years ago it was a very small handful of researchers mostly in Japan – I should mention there’s a lot of really great work going on in Japan as well. But ten years ago it was mostly a small number of Japanese research groups and now suddenly it’s very much exploded and all over the world there are so many people doing really great work. I work in the bio design lab at Harvard with Professor Connor Walsh, and in terms of our immediate interest, I think wearable robotics is the area that we’re really interested in, but there are a lot of other groups at Harvard as well who are doing really good research, ranging from biomimetic systems to different types of medical devices.

Ron Vanderkley: Thank you Dolan for your time. We’re excited to see what happens with the competition!

Donal Holland: Thank you very much for having me. We’d be delighted to hear from any of your listeners with any feedback or thoughts on the toolkit.

All audio interviews are transcribed and edited for clarity. However, we can not assume any responsibility for their accuracy.

ep.173: RoboThespian, with Will Jackson

Ron Vanderkley — Sat, 10 Jan 2015 07:47:43 +0000

Transcript below

In today’s podcast, Ron Vanderkley speaks with Will Jackson from Engineered Arts Limited about his team’s work making robot actors.

Engineered Arts was founded in October 2004 by Will Jackson, to produce mixed media installations for UK science centres and museums, many of which involved simple mechanical figures, animated by standard industrial controllers.

In early in 2005, the Company began work on the Mechanical Theatre for the Eden Project. This involved three figures, with storylines focused on genetic modification. Rather than designing another set of figures for this new commission, Engineered Arts decided to develop a generic programmable figure that would be used for the Mechanical Theatre, and the succession of similar commissions that would hopefully follow. The result was RoboThespian Mark 1 (RT1).

From thereon, Engineered Arts took a change of direction and now concentrates entirely on development and sales of an ever expanding range of humanoid and semi-humanoid robots featuring natural human-like movement and advanced social behaviours.

RoboThespian, now in its third version, is a life sized humanoid robot designed for human interaction in a public environment. It is fully interactive, multilingual, and user-friendly. Clients range from NASA’s Kennedy Space Centre through to Questacon, The National Science and Technology Centre in Australia. You can watch it in action in the video below.

Will Jackson
Will Jackson has a BA in 3D design from University of Brighton, UK and is the Founder of Engineered Arts Ltd.

Links:

ep.171: Robotics in Theatre, Film, and Television, with Grant Imahara and Richard McKenna

Ron Vanderkley — Sat, 13 Dec 2014 08:08:14 +0000

NEW: Transcript below.
In this episode, Ron Vanderkley speaks with Mythbusters‘ Grant Imahara, and Richard McKenna from The Creature Technology Company about robotics in the film, television and theatre industries.

Grant Imahara
Grant Imahara graduated from the University of Southern California with a degree in electrical engineering. It was shortly after that Imahara was hired as an engineer at LucasFilms and Industrial Light & Magic, building and operating a number of visual effects, models and robots for popular films/film series (such as Star Wars, Galaxy Quest, Jurassic Park, Terminator, The Matrix and AI: Artificial Intelligence). Imahara also built the Energizer Bunny, for the battery company’s commercials, Deadblow robot on BattleBots and Jeff Peterson from The Late Late Show with Craig Ferguson. Imahara is perhaps best known as a presenter on Discovery Channel’s MythBusters, and is often seen making robots or robotic rigs needed to aid in the testing of various myths. Imahara appeared at Supanova 2014 promoting his casting as Mr. Sulu in the popular professional web-series Star Trek: Continues.

http://youtu.be/iJv6LfI1yxk

Richard McKenna
Richard McKenna is Chief Engineer at The Creature Technology Company. He joined CTC in 2010 and has worked on all of the major projects since that time, including How to Train Your Dragon; King Kong and the Sochi Olympic Mascots. He has a Bachelor of Engineering (Hons) in Mechatronics, Robotics and Automation Engineering from Swinburne University and is certified as a “Chartered Professional Engineer” by Engineers Australia, registered on the National Professional Engineers Register (NPER). Prior to joining CTC, the majority of Richard’s time was spent in the defence industry, and he has also worked in special effects for film and television.

Links:

References:

Stanley “Stan” Winston was an American television and film special make-up effects creator. He was best known for his work in The Thing, the Terminator series, the first three Jurassic Park films.

Gerry Ryan of Jayco Caravans the largest recreational vehicles manufacturer in Australia, owner of Creature Technology, Global Creatures, (Gerry Ryan is listed in BRW Rich 200 list 2014: 148)

Dreamworks “HOW TO TRAIN YOUR DRAGON LIVE SPECTACULAR”, where dragons with 40 foot wingspans fly and breathe fire. This must-see arena spectacle breaks all the rules of traditional entertainment, immersing audiences into a magical and mythical world of Vikings and Dragons.

ep.168: Nylon Fishing Line Actuator, with Geoff Spinks

Ron Vanderkley — Fri, 31 Oct 2014 12:10:27 +0000

In this episode, Ron Vanderkley speaks with Professor Geoffrey Spink from Wollogong University about his team’s work on artificial muscles.

Professor Spink’s Australian led team are able to produce artificial muscles from nylon fishing line that can contract by about 50% of their original length, can generate forces that are over 100 times larger than our own muscle, and produce a mechanical power output of over five kilowatts per kilogram – similar to a jet engine. He tells us that “the advantage of using something as simple as fishing line or sewing thread is that they’re really cheap and they’re readily available… The tools we need to make the fishing line muscles are fishing line, a hair dryer, and an electric drill.”

The process is simply to attach one end of a piece of fishing line to an electric drill and the other end to a weight so that it provides a bit of tension. As the drill is switched on and starts rotating, the weight end of the fishing line is impaired from twisting, and eventually this leads to a phase called over-twisting, where the fishing line forms a coil. Finally, a blast of heat sets the shape and it is left to cool. To make the coil work as an artificial muscle, a little bit of tension is applied again, and then some more heat. The polymer fibers can be made very thin or very thick, with the amount of force generated increasing as the fibers get thicker.

The lifespan of the polymer cord muscles has been tested for one million cycles, and there was no degradation in performance.

Possible uses include artificial muscles for prosthetics, and the design of light weight humanoid robots. In development is a new lymphoedema compression sleeve, which will use these muscle fibres to gently massage the arm. The garment may prevent the build-up of lymphatic fluid and the discomfort that it causes. Woven into clothing fabric, this could create smart garments that are powered by body heat. If the wearer gets too hot, the muscles expand, the weave opens up and that releases some heat until the body has cooled down again. Similar concepts could be used to regulate the temperature in a building or greenhouse.

Geoff Spinks

Professor Geoff Spinks is an Australian Research Council Professorial Fellow. His research interests focus on new materials and nanotechnology and he specialises in the development and application of materials for artificial muscles.

Spinks has published over 160 journal articles, including 4 co-authored articles in Science magazine. He is the co-recipient of in excess of $35m in grant funding.

Spinks has worked closely with industry including a sabbatical leave with BHP Research and Allied Signal Inc. (USA) and collaborative projects funded through the ARC Linkage and CRC schemes. His current interests include new product development (such as the “Lymph Sleeve”) and new manufacturing methods (such as 3D printing) that use his artificial muscle materials.

Spinks has had a strong engagement with teaching across all levels of engineering materials and was co-founder of UoW’s bachelor degrees in Nanotechnology. He has supervised 22 PhD, 7 Masters and over 50 final year / honours students to completion.

Links:

ep.159: Computer-Assisted Surgery, with Karol Miller

Ron Vanderkley — Fri, 27 Jun 2014 15:47:16 +0000

In this episode Ron Vanderkley speaks to Prof. Karol Miller, Director of the Intelligent Systems for Medicine Lab at the University of Western Australia, about medical robotics and how doctors and patients perceive its role and use. Central to his work are mathematical models of soft tissue (brain, liver, etc.) that can be used for robot-assisted surgery by providing fast and accurate feedback. He also discusses the potential and challenges in making surgery fully autonomous. Finally, we hear about how his computational tools could be used to build transformer-like robots that can reassemble to adapt to new tasks.

Karol Miller
Karol Miller is a specialist in the area of modelling and computer simulation in biomechanics. He is the Winthrop Professor of Applied Mechanics at the University of Western Australia, the Deputy Head of the School of Mechanical and Chemical Engineering, Director of the Intelligent Systems for Medicine Lab and Director of the Computational Geomechanics Laboratory. His contributions include mathematical models of brain deformation behavior (the world’s most frequently cited) and close-to-real-time finite element and meshless algorithms.

Links:

ep.156: RoboRoach, with Bill Reith and Oliver Bendel

Ron Vanderkley — Sat, 17 May 2014 02:43:44 +0000

In this episode, Ron Vanderkley speaks with Bill Reith, an engineer at Backyard Brains. The company develops RoboRoach, the world’s first commercially available “cyborg”, which was successfully backed on KickStarter. Using their kit, you can briefly control the left/right movement of a cockroach from a smart phone using microstimulation of the antenna nerves. The RoboRoach is one of the many tools that Backyard Brains uses to teach students about neuroscience. Aware of the ethical questions their work raises, the company has put out a set of Ethical Statements. To explore the ethics further, AJung Moon speaks with Prof. Oliver Bendel, at the School of Business in Basel.

https://www.youtube.com/watch?v=63qwg7EBxbM

Bill Reith
Bill Reith is Engineer at Backyard Brains were he works on RoboRoach. He also uses the companies products as an educator at Maker Faires, university workshops, high school events, local bars, and TED talks.

Reith has a Bachelors degree in Bioengineering and Biomedical Engineering from the University of Michigan where he worked at the Neural Engineering Laboratory. He also worked at Maker Works, helping customers with their projects at Ann Arbor’s makerspace.

Oliver Bendel
Oliver Bendel is Professor of Information Systems and Ethics at the University of Applied Sciences and Arts Northwestern Switzerland, where he is active in the areas of e-learning, knowledge management, social media, mobile business, avatars and agents, information ethics, and machine ethics (http://www.maschinenethik.net/). His current areas of focus in machine ethics are chat bots, medical and care robots, advanced driver assistance systems and the combination of machine ethics and animal ethics.

Links:

ep.155: Zero Tillage Robotics, with Peter Corke

Ron Vanderkley — Fri, 02 May 2014 13:34:53 +0000

In this episode, Ron Vanderkley speaks with Professor Peter Corke from Queensland University of Technology, about the fast-tracking research that will see robots planting, weeding, maintaining and harvesting crops. The AgBot is a light-weight, golf buggy-sized robot that has been specifically designed to reduce the environmental impact of weeding. It can navigate wheat farms of around 4000 hectares using low-cost sensors, targeting weeds with spray while they are still very young plants. The Chief Investigators are Peter Corke, Ben Upcroft, Gordon Wyeth and Salah Sukkarieh (ACFR) with Partner Investigator Andrew Bate from SwarmFarm.

Peter Corke
Peter Corke joined Queensland University of Technology at the start of 2010 and is known for his research in vision-based robot control, field robotics and wireless sensor networks. He received a B.Eng and M.Eng.Sc. degrees, both in Electrical Engineering, and a PhD in Mechanical and Manufacturing Engineering, all from the University of Melbourne. Prior to QUT he was a senior principal research scientist at CSIRO where he founded the Autonomous Systems Laboratory, a 50-person team undertaking research in mining, ground, aerial and underwater robotics, as well as sensor networks. He subsequently led a major cross-organizational “capability platform” in wireless sensor networks.

Links:

If you liked this article, you may also be interested in:

See all the latest robotics news on Robohub, or sign up for our weekly newsletter.

From the Greenhouse to the Fields, with David Dorhout

Ron Vanderkley — Fri, 15 Nov 2013 18:56:35 +0000

Link to audio file (27:16)

In this episode, Ron Vanderkley speaks with David Dorhout from Iowa State University about his Agricultural Robots that include Prospero the robot farmer and Aquarius the greenhouse watering robot.

Today’s agricultural equipment has been designed around a single farmer controlling large machinery. This method has its drawbacks since farming decisions have to be made at the level of the field. Nature instead is chaotic and dynamic, soil nutrients and moisture change from foot to foot. A swarm of small robots like Prospero would have the ability to farm inch by inch, examining the soil before planting each seed and choosing the best variety for that spot. Ideally, this would maximize the productivity of each acre, allow less land to be converted to farm land, and ultimately feed more people.

Prospero is the working prototype of an Autonomous Micro Planter (AMP) that uses a combination of swarm and game theory to plant seeds at safe distances from one another.

Dorhout’s second agricultural robot Aquarius is a greenhouse robot that autonomously waters plants using its 30 gallon tank. The robot is programed using taping on the ground of the greenhouse.

David Dorhout
David Dorhout is a graduate of Iowa State University. He has always been interested in robotics and has 14 years of experience in the agriculture and biotech industry doing field and greenhouse discovery work. He is the founder of Dorhout R&D LLC which is a research and development business designing and building novel robotic systems and interactive consumer electronic devices.

Links:

FutureDairy, with Kendra Kerrisk

Ron Vanderkley — Fri, 18 Oct 2013 13:10:50 +0000

Link to audio file (38:51)

FutureDairy is an R&D development program to help Australian dairy farmers manage the challenges they are likely to face during the next 20 years. In this episode, Ron Vanderkley speaks with the project lead Kendra Kerrisk from the University of Sydney about robotic milking and herding.

As one of the big challenges is the availability of labor and associated lifestyle issues in the dairy industry, FutureDairy’s focus is on automatic milking systems. While robotic milking technology is now in wide use overseas, there’s less experience with automatic milking in grazing-based farming systems such as in Australia. The video below summarizes some of the results from the DeLaval pilot farm in Australia.

With increasing numbers of Australian dairy cows now being milked by robots, researchers are looking at a range of exciting ways to use robots on the farm. One that has already shown promise is the use of robots (UGV) to herd cattle from the paddock to the dairy. In the videos below you can see two trials with the Shrimp rover hearding cows in Australian farms.

FutureDairy 3 is sponsored by Dairy Australia, the University of Sydney, the NSW Department of Primary Industries and DeLaval.

Kendra Kerrisk
Kendra Kerrisk is Faculty of Veterinary Science at the The University of Sydney in Australia. Kerrisk developed a strong interest for the Dairy Industry whilst conducting undergraduate and post-graduate studies at Massey University in New Zealand and has remained a leader in the field since then. During her PhD at the University of Melbourne she studied Peri-parturient Management for Large Dairy Herds using Controlled Breeding Programmes. After her PhD, she worked for Dexcel (formerly Dairying Research Corporation) in New Zealand on the world’s first pasture-based Automatic Milking System research farm. Through her academic life, she has contributed significantly to the national and international knowledge regarding application of Automatic Milking Systems (AMS) with pasture-based dairying. One of the highlights of the work conducted within FutureDairy, a project which she leads, has been the co-development of the world’s first Robotic Rotary (Automatic Milking Rotary, DeLaval AMRTM). This internationally recognized work will increase the feasibility of robotic milking for large dairy herds that are more common within the Australian and New Zealand industries.

Links: