In a surprise move today, Toyota held a press conference (see video below) announcing a substantial investment in robotics and AI research to develop “advanced driving support” technology, with former Program Manager of DARPA’s DRC Gill Pratt directing the overall project as Executive Technical Advisor. Toyota will allocate USD$50M over the next five years in a partnership with MIT’s CSAIL (headed by Daniela Rus) and Stanford’s SAIL (headed by Fei-Fei Li) to develop research facilities in Stanford and Cambridge.

Two main themes from the press conference underline the tone of this effort:

First,  Toyota doesn’t refer to autonomous cars but to ‘advanced driving support’ . And this makes sense because, unlike Google (with its fully autonomous car), Toyota’s origins are in car manufacturing and not in software development. Car manufacturers have for some time been introducing ever rising levels of autonomy in their cars and Toyota is following that vector. There are legal advantages to having the driving be ultimately responsible for the car’s movements and therefor any possible accidents; when a vehicle doesn’t respond as expected, a costly product liability suit could follow, as Toyota well knows. Thus what appears at one level to be a more conservative approach to autonomy on the road may in essence be a way to bypass potential legal nightmares and bring autonomous driving technology to the market sooner.

Kiyotaka Ise, Toyota Motor Corporation Senior Managing Officer, repeatedly stated hybrid cars as an example of a specific technology that Toyota pioneered (along with Honda) and brought to market even while it was borderline unprofitable due to its large cost at the time. This offers a possible hint of what may follow.

The second interesting point made clear during Toyota’s Q&A is that this project is only part of the auto manufacturer’s investment in the field. $50M is a lot of money, but for a profitable car maker the size of Toyota, it isn’t much. It looks like Toyota wants to establish a strong presence in the US by collaborating with two of the best research centers in the world (the budget will be equally split between the two universities) … but being a truly global company, it is developing other projects elsewhere as well. It’s worth mentioning that almost all large car manufacturers have similar research projects, and they usually collaborate with research institutes, universities or with each other.

In this respect, Toyota’s decision to bring on Pratt as the project lead appears astute. Pratt’s high visibility role as Program Manager of DARPA’s DRC has left him with significant experience managing and promoting large, big-budget projects, and has provided him with deep connections into US research labs.

You can read Toyota’s press release here for all the details. Automaton has also an excellent coverage of the story (see Toyota Announces Major Push Into AI and Robotics, Wants Cars That Never Crash) and says they will be releasing more in-depth stories about the deal early next week. You can watch the full press conference below:

When Google bought Boston Dynamics last December, the news made headlines, but it was not the first time the Internet giant has invested in DARPA-funded robotics. As part of Robohub’s Big Deals series, we asked Gill Pratt, Program Manager of DARPA’s Defense Sciences Office, to shed some light on what DARPA thinks about Google’s robotics acquisitions , and what it might mean to the robotics and open source communities.

Congratulations, Gill, on the DARPA Robotics Trials – what was your personal assessment of the event?

I was very pleased with how well the teams performed, and in particular that the hardware for the majority of the teams worked so reliably. In general, the teams did slightly better than we had expected and this was probably the result of a lot of very careful preparation. The infrastructure that we set up also worked reliably in the trials. Still, the DRC Trials were really just a beginning and showed us the potential of what can be done in disaster scenarios with robots.

Cloud computing was an important facet of the DARPA Robotics Challenge and especially the Virtual Robotics Challenge … How do you see cloud computing helping the development of robots?

The importance of cloud computing for Robotics falls into two different categories. The first has to do with simulation environments. We used cloud computing for the first time with the Virtual Robotics Challenge, which we had back in June as a precursor to the DRC Trials, to do real-time simulation of tasks in unstructured environments. We rented a whole lot of cloud computing space, we ran a very good simulator developed for DARPA by the Open Source Robotics Foundation, and we adjusted the latency – the time delay – so that it would be the same for all the teams, even though they were located in different parts of the world. It was a tremendous enabler to run a real-time competition based on simulation that also involves human-robot collaboration. Doing so required that the simulation run in real-time. That was a big first and we felt it was very successful.

Another important aspect of cloud computing is something that we have not used to date, but I think has a lot of potential for the future: the ability to have a robot’s computing and data hosted remotely. There’s a lot of potential for this, in particular for doing perception. Forty percent of the human brain is used for perception, and perception is one of the most difficult tasks for an autonomous robot. It’s very difficult to fit a computer with the size, weight, and power that you need to achieve really good perception onto a robot: the computer just gets too large, it consumes too much power, and it weighs too much. However, if you have access to cloud resources with lots of data and lots of computing cycles, they can help move that burden off of the physical robot.

Perhaps most exciting for the future of cloud computing in robotics is that when one robot learns how to perceive something, or learns how to do a particular task, that learning can be instantly shared with other robots. This sharing could have a catalytic effect on the capabilities of robots, particularly in structured environments.

This is really an up and coming field. How likely are we to see commercial interest come out of cloud computing for robotics?

If you look at the research being done we see a lot of possibilities with cloud computing for robots. In the commercial world, I think we are going to see it applied first to structured environments where there are lots of human artifacts — things like doors, stairs, furniture, tools — and where a large database of stored examples of these artifacts can help the recognition problem.

The Atlas robot, created by Bostron Dynamics and DARPA, was used by several teams in the DRC Trials. Source: DARPA

This is an interview for our Big Deals series, so we’re wondering, what does DARPA think about Google buying up some of the technologies that DARPA has spent years building? We’re thinking of the cars, the humanoids …

We are thrilled to see commercial interest. It’s one of the signs of success for the investments that we have made in future technology.

We think that it takes commercial investment to drive down the cost of technology. A very well known example is the cellphone. Cellphones have microprocessors in them that many years ago began with investments by DARPA’s Microsystems Technology Office and others. Cell phones have inertial measurement units to figure out the tilt of the cellphone. They have displays; they have GPS receivers; they have radio devices that work at low power. I can point to any number of DARPA investments that helped start those technologies, and of course cellphones now talk to the Internet, which is probably the best known output of all of DARPA’s investments. But if the Department of Defense wanted to produce a cellphone without the commercial world having picked it up and turned it into a product that billions of people use, it would cost many orders of magnitude more than it does now. It’s really because of the commercial world that we’ve seen the price of cellphones go down.

But DARPA is not only interested in seeding technologies with these early investments; we are also committed to solving real problems for national security. In fact, we have a variety of programs that use cellphones. One of these is called TransApps, where soldiers use smart phones to plan and carry out missions and layer and share data between teams. This is an example where investment by the commercial world made possible a national security capability that otherwise might not be feasible because of excessive cost.

By analogy, what’s going on in robotics is that we’ve seen some interest in the commercial world in taking this to the next step. And they have resources far beyond what DARPA has to further develop a technology and, most importantly, to drive down cost.

So these big deals are a good thing for DARPA?

Yes, they are a wonderful thing for DARPA and, more broadly, for our nation.

Do you think it’s a good thing for the field of robotics in general?

Absolutely, but I understand why there’s some anxiety.  It’s true that, for a very short time, some talented people in the field will be working for a commercial firm rather than for DARPA. However, that effect is transient and very small compared to the much more important thing, which is that these corporate investments both accelerate the field and show that it is going to be real.

This anxiety is a lot like the kind you get when your kids leave home. It’s sad, and for a while you’re going to miss them, but wait a while and then the grandkids come. The important thing is for Robotics to become a field that is not only about prototypes or lab demonstrations, but a field where significant resources are invested to make products that really improve people’s lives. We saw it with the Roomba vacuum cleaner, but we want to go beyond that. These kinds of commercial investments will make the field far more attractive to startups and to students who are planning their careers. The net effect is strongly positive.

How do you think this corporate interest in robotics will impact the open source community?

Open source is very important. It allows us to develop pre-competitive catalysts to help the whole field move forward.

In the integrated circuits world it was recognized that you couldn’t build experimental integrated circuits as easily as you could wire up a circuit with discrete components like capacitors and resistors and transistors that you solder by hand. As a result, people working in the area at that time needed to develop a simulator to allow for the experimentation of how a chip would work without actually having to make the chip. The idea was to do the design and the debug cycle on the computer in simulation instead of having to do it on the bench: when you were done you could press a button and have confidence that the chip would actually work. The open-source simulator SPICE filled that niche.

The very same kind of thing is true in the robotics field. The key is to recognize that simulation is really a pre-competitive technology now.  We don’t need to have proprietary closed simulation systems and keep reinventing the simulator over and over again. We’ve done it enough times now to know that it would be useful to apply open source. In the same way, I think that an operating system is precompetitive – it has been done enough times that a new company should not try to distinguish itself based on a proprietary version of an OS.

Rather, let’s just move on to the next thing where the real innovation is going to occur. DARPA funded the Open Source Robotics Foundation to develop an open source simulator for the DARPA Robotics Challenge because we think that it will catalyze the whole field to everyone’s advantage.

With the purchase of Boston Dynamics, some headlines were saying that Google bought military robotics. Is this an accurate assessment?

I think it’s very inaccurate. It’s based on the confusion – and the conflation – of two aspects of robotics: remote bodies and remote brains. A robot allows you to have a mechanical body take the place of a human being; for example, a robot can work in a dangerous environment while a human supervisor is in a safe place. And a robot also allows you to perform some of the function of the human brain autonomously in a remote location. When you conflate bodies and brains, and say that there is only one idea here, it’s like saying, “The robot looks like person, so it must be as intelligent as a person, and therefore I should be scared of all robots, particularly those funded by the Department of Defense.”

All of the robots that I’m aware of, except for the ones that actually have weapons or specialized sensors on them, are generic. They are neither military nor commercial. They are just robots that have mobility and perhaps some manipulation, but there are not classed to one type or the other. And they are all almost completely empty-headed: their bodies may look like ours, but their heads are almost totally empty.

In terms of autonomy, the technology has such a long way to go that the fears that have been generated are way out of proportion to the state of the art. At the DRC Trials, for example, we had people supervising the robots in pretty tight loops, telling them what to do. While autonomy may be able to handle very simple environments, close human supervision is going to be needed in unstructured environments for some time.

What do you say to people who are uncomfortable with the use of military robots in a commercial context?

I think there are legitimate concerns, but from a rational point of view it’s important to distinguish between a weapon, which none of the robots we are talking about are, and a machine that makes us feel scared because it looks like us but it’s hard to understand exactly how it works. Of course, science fiction has stoked that fear with movies like Terminator and so forth, but that reaction actually goes back a long, long way to stories like Frankenstein – anything that sort of looks like a person but is hard to understand will generate fear. It’s an emotional reaction. However, from a rational perspective, if you consider how little autonomy these real machines actually have, you realize that the worry is way out of proportion to reality.

What are some of the benefits that could come out of combining Google’s expertise in data with hardware from Boston Dynamics?

I don’t know the specific applications that Google is thinking of – it’s their prerogative to discuss them or not, and they haven’t told DARPA.

Generally speaking, when you think about data and robotics, there are a lot of exciting things that could happen. For example, you could do perception better by using a lot of data and sophisticated search. This is not necessarily the kind of search we think of when we type words into our browsers, but the kind that involves images, where a robot could say, “Okay, I see something and because of many prior examples I know what to do,” and share that data in a useful way.

Would voice commands play into this dynamic, in addition to visual search? Would it make it easier for humans to interface with robots?

I’m not an expert on voice, so I don’t want to speculate, but I know that you can train better when you have a lot of data. Voice and natural language recognition have the potential to help with planning and perception, and I think they could also facilitate more intuitive interfaces for sharing information between robots and humans.

Team SCHAFT raises the arms of its S-One robot in victory after successfully completing the Climb Industrial Ladder task at the DRC Trials. SCHAFT won that task and three others, and scored the most points of any team at the event. Source: DARPA.

What new robotic technologies will DARPA be pushing that might appeal to big companies that are not traditionally robotics focused?

We’re really interested in cooperation directly between robots and between human beings and machines. We are also interested in dealing with dynamic environments. What I would like to do for the future of the DRC, and what I think that DARPA is going to push even beyond that, is to answer questions like “What if the communication is intermittent?” and “What if the environment is beyond a set of familiar things that I’ve seen before?” … What do we do if we’re actually outdoors? How do we handle this piece of vegetation that’s in front of us? Do we go through it or around it? If we go through it, do we move the branches out of the way?

We also still have a lot of research to do to figure out how to apply cloud computing when the environment is unstructured, like it would be in disasters or outdoor scenarios.  And if the communications are intermittent, we have to figure out how to cache enough of the intelligence on the robot itself so that it can continue to do what it needs to even when it’s disconnected from the cloud. Of course, disconnected operation may be at a reduced level of effectiveness, but we don’t want the robot to stop altogether when communications go down for a little while. We want the robot to have at least some ability to complete the tasks it was given to do.

How do you actually put behavior in the cloud? How do you take a data driven approach to behavior? I think that’s very much an unknown thing right now. Data driven approaches to perception are easier to understand, but we are going to try to take the next step and look at behavior as well.

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

• Snowden BEAMs into TED: How robotic telepresence disrupts borders
• Google’s robot and artificial intelligence acquisitions are anything but scary
• Will companies like Google and Amazon usher in a new era for robotics?
• The robots behind your Internet shopping addiction: How your wallet is driving retail automation
• What do Google, Amazon, IBM, Apple, Microsoft and all the airlines and car companies want?
• Robots Podcast: Privacy, Google, and big deals
• Robots, soldiers and cyborgs: The future of warfare

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

Based on a media rountable discussion with DRC Program Manager Gill Pratt and CEO of the Open Source Robotics Foundation Brian Gerkey.

The Fukushima Daiichi Nuclear disaster was a wake-up call to the robotics community. In Japan, many asked why a country known for its cutting-edge robotics sector was unable to respond to the emergency. Worldwide, robotics experts pointed to the event as a real-world test of what robots can and cannot do.

Whether man-made or natural — or like Fukushima, a combination of the two — major catastrophic events, while rare, are becoming increasingly costly as human populations worldwide move to urban areas. This is why, in an effort to spur the development of agile humanoid first-responders, the US Department of Defense’s strategic plan identifies disaster response as a priority area, and why it is funnelling tens of millions of dollars into the DARPA Robotics Challenge.

DRC Program Manager Dr. Gill Pratt envisions a future where tele-operated robots can be quickly deployed by disaster-response personnel with minimal robotics training: “In Fukushima, during the first 24 hours, the explosions would not have occurred if it would have been possible to vent the reactors,” said Pratt in a media round-table last month. “Human beings tried to do this but had to turn around [because of the radiation]. Our hope is to develop machines that can actually intervene and help make a disaster [like this one] less severe.” Estimates of the total economic loss due to the Fukushima disaster range from $250B – $500B US.

Why a humanoid?
Though having a robot that can safely navigate dangerous places like the inside of a nuclear plant is no-doubt useful, one has to question whether the extreme cost and design challenge of developing a humanoid that is capable of doing so is worth the effort. Yet as Pratt points out, a disaster scenario that is catastrophic to humans is also most likely to occur in an environment that has been engineered for humans, and what kind of robot would be more capable of navigating in a human environment and using the tools found there than a humanoid?

As DARPA project leaders were designing parameters for the Robotics Challenge, they consulted with first-response experts — including the Nuclear Regulatory Commission — and identified key capability requirements that indeed seem to point to a humanoid solution: the ability to go through doors, climb steep stairs, walk on rough terrain, and the ability to use whatever tools are oh hand, such as a utility vehicle or a screw driver.

“Our belief is that we will never know where the next disaster is going to hit and when it will be, and so we have to be prepared to use whatever tools are on hand,” said Pratt. “And at least for the foreseeable future, those are tools for people to use, and so we need to adapt our machines to use those tools as quickly as possible.”

Designing a roadmap for success
DARPA does not define the success of its programs through a profit and loss model, and its goal with the DARPA Robotics Challenge is not to bring a humanoid first-responder to every fire station in the country. As a government agency, DARPA focuses on bringing high-risk technologies to the point of prototype, and then handing these off to the military or commercial sector for development. It invests in challenges such as the DRC in order to spur radical innovation, and short-term technical failure is OK as long as the long-term potential payoff is high.

But despite these lofty goals, the question remains: what’s the roadmap for designing and overseeing a major robotics challenge such as this one? How does one gauge success when the potential payoff may be decades or more away?

DARPA’s previous autonomous car challenges offer some lessons, and one of them is that there isn’t always a winner.  Pratt admits that success for the DRC isn’t guaranteed, but says that the goal of the competition isn’t so much to declare a winner as it is to continually nudge the competitors towards greater achievements, while at the same time being realistic about what is actually doable. As a result, DARPA will be reevaluating the project’s milestones and adjusting the difficulty of the tasks in consultation with the participants along the way.

Previous autonomous car competitions are different from the current challenge in one major aspect, however: the older competitions were based on automotive technology, which already had a significant infrastructure in place. The robotics industry, by comparison, is still in its infancy, and so a great many hurdles exist for competitors, both at the software level and at the hardware level, and in terms of integration. Pratt has thus broken the competition down into three stages and four separate competitor tracks in an attempt to structure what is otherwise an extremely open-ended problem. (You can find a good explanation of how the competition breaks down on DARPA Robotics Challenge Fact Sheet.)

The first stage of the DRC — the Virtual Robotics Challenge (VRC) — offers particular insight into DARPA’s strategy for success. The VRC, for which the results were announced earlier this month, was a cloud-based competition designed to test software, and its goal was to evaluate the teams’ robot perception, manipulation and locomotion in a virtual environment as a first step before moving on to the physical trials, which will take place in December of 2013. The big idea here is that teams will be able to try out their strategies in simulation first, before investing in hardware. “The reason we had this virtual challenge is that we wanted to open the contest to teams whose strength is primarily software,” said Pratt.

DARPA’s simulator models real-world physics and behaviours with incredibly high accuracy. Brian Gerkey, CEO of the Open Source Robotics Foundation (OSFR), the organization behind the VRC simulator, says that their goal is for the simulation to be the best possible stand-in for the real robot, so that teams that have done well in the VRC should be able to run their software virtually unchanged on the physical robot during the trials. And yet run to such a complex simulation — and to play that simulation back at interactive rates — takes a huge amount of computer power.

“One of the key developments over the past several years has been the advancement of cloud computing resources,” explained Gerkey. “It’s now possible to rent access to cloud computing resources so that you can have modest computer at your end and have the simulation run somewhere else. That’s a key aspect of this project and it’s really required in order to run the competition.”

The seven winning teams from the VRC challenge each received an Atlas furnished by Boston Dynamics for use in the DRC trials, where they will compete with other funded and unfunded teams that have been developing their own hardware (see DARPA’s Meet the Teams video to get a sense of what Atlas will be up against). They now have until late December to transfer their software algorithms to the Atlas hardware shell, and practice running a real machine in real environment. The physical trials should be a good demonstrator of the success of the simulator approach to competition design.

Building a legacy
DARPA takes a long view to technology development, starting with identifying a major national security need (in this case: disaster response), then determining the specific technologies and innovations needed to address that need (in this case: easily deployed agile humanoid robots), and finally working to support the fundamental breakthroughs required to make these technologies possible. DARPA’s disaster challenge is thus about creating infrastructure to support long-term software and hardware development in the robotics community; it’s also about attracting bright minds to solve some of the world’s most pressing problems.

One of the legacies of the VRC is that all the software code for the simulator is being distributed open source, and so, as Gerkey points out, there is no cost to reuse it later.

“The history of software development has shown that this is the best possible, most efficient development model for building the common platform that a whole community comes to rely upon,” says Gerkey. “The simulation work that we’re doing is really part of a broader effort to build a common ecosystem of software tools and libraries that everyone can use. There are computers that are powerful enough to run this simulation that you can rent for on the order of a couple of dollars an hour.”

The hope is that, ultimately, the open-source VRC simulator will help speed up the development of robots in general by lowering the barrier of entry and that this, in turn, will lead to more capable robots – perhaps ones capable enough to be of service in a Fukushima-like disaster. Says Gerkey: “Once we can run this very realistic simulation in a competition, we can then try this in a classroom scenario. You don’t have to outfit your classroom with expensive powerful computers.”

Pratt agrees: “Right now we have a few really talented designers and a few companies that are beginning to be strong in the robotics field. We would like to see those numbers go way up, and we believe that the way to do it is to give people a tool to develop these machines without having to incur the cost of first building them.”

Fostering international cooperation
Fukushima is used as an example to motivate the need for robotic platforms in dangerous environments, and the VRC is being promoted as an internationally beneficial endeavour, so I asked Pratt what specifically is being done to foster international cooperation, and whether there was a possibility that DARPA would work with other countries on future challenges.

Pratt answered that a key measure for success in this project is engagement with the international community. “All the code that we’re writing is distributed under the Apache 2 licence which is a very permissive open source license that allows anyone in the world to do anything that they want with this software. We would love for the challenge (once we are done with it in the next year and a half) to move around the world, and we have been in discussions with our counterparts in Japan and the EU about the possibility of this becoming a rotating contest. We think that [making societies more resilient to disasters] is a need that the whole world has, and so and I’m spending quite a lot of my time having lots of discussions about what the future of this is going to be once we are done … I personally think that the international part of this is a key way to measure how well we have done, so if we’re successful in having the DRC continue on somewhere else in the world, then it will be clear that we have done well.”

Though Pratt hesitates to predict how long it will be before we have robots capable of helping out in situations such as the one in Fukushima, he is convinced that DARPA’s Robotic Challenge will result in some major developments for the field:

***

More info:

DARPA Robotics Challenge Website
DARPA Robotics Challenge Fact Sheet

Look out for DRC trial coverage in December!

The first competition of the DARPA Robotics Challenge (DRC) was a software simulation of humanoid rescue work. The 6 best teams were to be allocated ATLAS robots for the next stage of the challenge, but there’s been an interesting change. Gill Pratt, the DRC program manager, and Nate Koenig, the CTO of OSRF, will be on reddit from 1:00 – 2:00 PM ET today (June 27) to discuss the results of the Virtual Robotics Challenge (VRC) and the future of the DARPA Simulator with the community.

In the Virtual Robotics Challenge (VRC), teams competed in a simulated suburban obstacle course. Twenty-six teams from eight countries qualified for the VRC, which ran from June 17-21, 2013. Competing teams applied software of their own design to a simulated robot in an attempt to complete a series of tasks that are prerequisites for the next stages of the grand challenge.

Gill Pratt: “The reason we wanted to have this virtual challenge is that we wanted to open the contest to teams that were strong not just in both building hardware for robots and programming them with software, but to include a wide variety of teams including those that had experience with just software. So for those teams whose strength is primarily software, we are having this contest in simulation. They will essentially be programming and interacting with a robot that we will provide for the people who will do the best in the contest, which is a robot called Atlas that was designed by Boston Dynamics. The government, at its expense, will give up to six teams who do the best in the VRC this robot to use in the [final] event itself.”

While DARPA had only allocated resources for the top six teams to move through, a combination of good sportsmanship and generosity will see members of the top nine teams move forward: