Hardware experiments on motion planning for the the ballbot Rezero using direct transcription. Source: ADRLab ETH/youtube

When you walk across a room or down a path, your brain is making thousands of decisions on how best to move. For example, how best to use your weight, scanning for any obstacles or uneven surfaces, and how rigid (or soft) your limbs and joints should be. Teaching a robot to conduct the same decision-making process is ongoing in robotics, and a team from ADRL, ETH Zurich and NCCR Robotics is studying existing direct transcription methods for trajectory optimization applied to robot motion planning.

Rezero, the dancing ballbot. Source: ETH Zurich.

Using a method of control called direct transcription (where complex mathematical problems are broken down into smaller problems and each solved individually), the team uses direct transcription to enable an unstable ball-balancing robot to perform a series of tasks with increasing complexity. The common issue with direct optimisation methods, which are used to allow the robots to obtain more natural movements, is that they require computers to continuously run multiple algorithms at once, meaning that planning a path in real time, like the human brain does, has not yet been achieved. Simply put, the computers working online with a robot are nowhere near as fast, efficient and robust as your brain, and that’s before considering how heavy such a computer might need to be, or how much bandwidth this communication requires.

First, by using computer models, the team tested the unstable ball balancing robot (see video below) with three variations of a simple task where the robot had to move from one location to another while avoiding fixed obstacles. By allowing the robot to use the best solution it found for previous tasks, coupled with a feedback controller to stabilise the system, the simulated robot was able to find a path through two obstacles in under a second. When using the real robot, the same paths and trajectories were followed, with the robot reaching the planned destination safely and in the same period of time as the virtual robot, thus validating the hypothesis.

The speed with which the robot is able to assess its scenario and follow a path that it has decided for itself without falling is a positive step forward that can be transported onto more complex robots (such as quadrupedal robots) in more uneven environments.

If a quadrupedal robot, such as HyQ or StarlETH, are able to understand obstacles in its path and successfully avoid or modify a movement to accommodate, such as softening joints when walking over rocks, then robots have made one step further towards regularly being sent to disaster zones to locate victims and save more lives.

Designing robots that can function outside of controlled, factory environments isn’t easy. When those environments have people in them, they are especially unpredictable and difficult for a robot to navigate.

Enter Rezero: a compact ballbot that can fluidly drive in any direction without prior orientation. Similar in size and shape to a human, Rezero can stabilize, instantly accelerate, avoid obstacles and turn around its own axis. It can also be programmed to follow a path, or a person at a constant distance, and it can even recover from jostles, making it ideal for navigating crowded environments like tradeshows, restaurants, offices and airports.

Possible Rezero applications: personal mobile medical device, telepresence tour guide, personal transportation.

It’s easy to imagine it one day being used to unobtrusively serve drinks at a cocktail event, as a personal mobile device for intravenous equipment at a hospital, or topped with a telepresence device and acting as a museum tour guide. Rezero could even one day rival the Segway as a personal transportation device.

A master of fluidity

Like an inverted pendulum, Rezero (left) balances on a spherical ball; three omniwheels (right) transfer motor action to the ball below.

Rezero’s agility comes from a propulsion and stabilization mechanism that keeps the device balanced on top of a single spherical wheel, like an inverted pendulum; it’s kept from falling over by constantly moving its contact point below its center of gravity.

A sensor called an Inertial Measurement Unit (IMU) measures Rezero’s tilt angle, which is used by an onboard stabilizing controller to compute what kind of motor action is required to keep the system stable. Instructions are then sent to electric motors that actuate three special wheels (called omniwheels), which in turn transfer forces to the sphere below: forces are applied in the rolling direction, but allow for free motion in the lateral direction; opposing each wheel is an arrestor that keeps the ball tightly in place. This process of measuring, computing, and actuating happens 160 times per second, and the underlying controller is key to achieving robust and dynamic movement.

Its ability to quickly respond to a changing environment is what makes Rezero so suitable for human environments, and according to Rezero’s creators, achieving this robustness was their most challenging task. Rezero’s controller is derived from a three-dimensional mathematical model of the system, and can combine complex and fluid movements around all axes. Compared to other ballbots, which can only lean a few degrees, Rezero can tilt up to 17 degrees as it leans into a curve at high speeds.

But getting there wasn’t easy. Rezero was developed in 2009/2010 as an educational project at the Autonomous Systems Lab at ETH Zurich,  and the 13 undergraduate students on the team were responsible for everything from concept to engineering, manufacturing and budgeting.

According to project leader Péter Fankhauser:

“It was the moment we got Rezero to balance for the first time that struck me the most. We got to know the system and work with it and were able to observe how its capabilities improved every day.”

Michael Neunert, his fellow founding team member agrees:

“The development phase was a very intense time. We only had nine months to sketch out our ideas and build Rezero as we see it today.”

Rezero continues to evolve even though some of the original team members have already graduated and moved on. The current team has just finished a new version of the omniwheel, which they have made available to other research groups. They are now working to improve Rezero’s robustness so that it can better manage obstacles in an everyday environment.

Key infos

Robot name: Rezero
Mechanical engineering: Simon Dössegger, Péter Fankhauser, Corsin Gwerder, Jonathan Hüssy, Jérôme Käser, Thomas Kammermann, Lukas Limacher, Michael Neunert
Electrical engineering: Samuel Schüller, Swen Wigert
Industrial design: Jessica Baumgartner, Tu-Van Giang, Claudio Maestretti
Project supervisor: Prof. Dr. Roland Siegwart, Director of the ASL at ETH Zurich
Websites: http://rezero.ethz.ch, TED Talk
Status: Ongoing research project

Photo credits: Rezero in Motion (first photo): Karl-Heinz Hug, Ringier AG; Rezero in the Studio (third photo): Gerhard Born, Ringier AG.

Note: This post is part of our Swiss Robots Series. If you’d like to submit a robot to this series, or to a series for another country, please get in touch at info[@]robohub.org.

My colleagues at the Autonomous Systems Christmas Lab at ETH Zurich have created an awesome Christmas video featuring some of their cool robots:

Mark Hoepfinger writes:

Dear All,

The 2012 Christmas Video from the Autonomous Systems Lab, ETH Zurich. The video illustrates the process of autonomous Christmas present distribution. Starring: Rezero, Firefly, StarlETH and the robotic elevator.

Merry Christmas and a happy New Year!!!
The ASL Team

For more holiday videos, have a look at the ROBOTS Podcast’s Annual Holiday Season Video Special Youtube Playlists. To submit your own, just send us a message at christmas@robotspodcast.com!