Péter Fankhauser

PhD Student - Robotics, ETH Zurich
Zurich, Switzerland

About Péter

Bio

Péter Fankhauser (born in 1987) is a PhD student at ETH Zurich in robotics and mechatronics. He is currently focusing on the control of legged robotic systems. Péter graduated with a Master's degree in Robotics, Systems and Control from ETH Zurich in 2012. During his studies, he took courses in entrepreneurship at the Massachusetts Institute of Technology and in robotics at the Tohoku University in Japan. For his Bachelor's degree, he led a team of thirteen students in the development of a unique robot. The robot Rezero is able to balance and drive on a single ball and thus moves in a very organic and elegant way while performing different tasks. Setting new standards in capabilities and performance of similar robots, his team was invited to conferences worldwide and Péter has given talks among others at Zurich Minds 2010 and TED Global 2011. In the future, Péter would like to combine his passion for mechatronic systems and entrepreneurship by creating innovative robotic products for the consumer market.

TED Conference

TEDGlobal 2011

Areas of Expertise

Robotics, Systems and Control, Product Development and Innovation

I'm passionate about

great products and technologies.

Universities

ETH Zurich

Favorite talks

Comments & conversations

672fcc0439ad253f775198bcaba25f15ecabb919 50x50
Péter Fankhauser
Posted over 2 years ago
Péter Fankhauser: Meet Rezero, the dancing ballbot
The main issue with dirt and dust is that the ball loses its "grip" to the floor. This is not a problem for normal maneuvers such as following a person, but in more dynamic movements this could bring the ball to slip and the robot eventually to lose its stability. We usually clean the ball after a full day of operation of Rezero. This is done by removing the ball and rinsing it with water or using a wet towel or similar. This process is very easy and quick. So far (after 18 month), there was no need to clean the wheels. Of course, an automatic cleansing feature would be desirable in the future.
672fcc0439ad253f775198bcaba25f15ecabb919 50x50
Péter Fankhauser
Posted over 2 years ago
Péter Fankhauser: Meet Rezero, the dancing ballbot
Thanks for sharing your thoughts. We use omni directional wheels to minimize the friction when the ball's surface is moving in the axial direction of the wheel. You can see this for example nicely in the video at 4:33. These wheels consist of small rollers on the circumference of the wheel and allow us to drive Rezero with very little friction loss. We have thought about different mechanisms for driving the ball. The mouse ball is usually read by two rollers positioned on the "equator" of the ball. This eliminates the capability of turning the ball around the vertical axis. Also, this needs extra passive contact points on top of the sphere to keep the robot on top of the ball which increase the friction.
672fcc0439ad253f775198bcaba25f15ecabb919 50x50
Péter Fankhauser
Posted over 2 years ago
Péter Fankhauser: Meet Rezero, the dancing ballbot
I consider normal usage when demoing Rezero for example at a conference. When demoing, we drive Rezero around back and forth and in circles mostly in a fix area (a lot of acceleration and deceleration). The max. speed in demos is usually little more than a human while walking. The terrain is mostly flat ground such as concrete, carpet, marble, wood or similar. I assume that the suction mechanism uses the most power in your robot vacuum. Also, the battery of Rezero is probably much bigger then the Roomba's. I hope this helps, even though I cannot tell you an equivalent distance to the Roomba vacuum.
672fcc0439ad253f775198bcaba25f15ecabb919 50x50
Péter Fankhauser
Posted over 2 years ago
Péter Fankhauser: Meet Rezero, the dancing ballbot
Hi Sloven, Thank you for your questions. We drove Rezero up on a ramp for handicapped people. This went without problems even though the robot/control didn't know about the ramp. We have videos on our website and YouTube where you can see this. You can push Rezero fairly hard (much harder then I did on the TED stage). However, if you kick it like BigDog you can make the robot fall. The big ring around Rezero protects in such a case the mechanics and electronics. All the control is done on the robot itself and an external computer is used only for remote input. Cheers!