The challenge of robotic locomotion

0

Legs or wheels – four or two? These are fundamental design choices if you are building a robot. For drones, there are other issues starting with the issue of staying aloft. There is, however, room for clever design to promote versatility. Anyone interested in reinventing the wheel?

Boston Dynamics was with Hyundai, which now owns 80%, at CES where they put Spot through its paces and followed with some less familiar robots:

https://www.youtube.com/watch?v=i0nJQr6GANY

Spot is a fairly familiar robot by now, but we’ve never met Mobed, a small, four-wheeled rig with a mind of its own. It doesn’t look that impressive at first, but when you scale it up, you have a self-contained pod.

These platform-based designs rely on four wheels, which limits the range of terrain they can handle. Having just two wheels can be more versatile as Ascento Pro demonstrates:

The advantage of Ascento Pro is speed, up to 12km/h, but what can it be used for? Transport yes, but also surveillance and inspection, where the main problem limiting its usefulness is its size.

Researchers at UC Berkeley’s HiPeRLab have developed a new quadcopter that can change shape in mid-flight using free-spinning hinges that allow the vehicle’s arms to fold downward by reducing or reversing forces of thrust. As this video shows, the configuration change allows the vehicle to traverse small passages, perch on overhead wires, and even perform grabbing tasks:

Finally, let’s take a look at some weird methods of locomotion:

These are from Carnegie Mellon where each semester, new students in the Robotomechanics Lab are challenged to do the dumbest step possible on any robot in the lab:

Stupid, yes, but also thought-provoking and some might even merit further investigation. A robot that moved by sliding could work well. The problem is that it is still far from easy to consider the mechanism of the robot as a mechanism without any of the intelligences that computer control can bring. To an engineer, two wheels are pretty useless without a human pilot, but to a computer engineer, they can be independent and more versatile than the basic mechanism suggests. Without computer control two wheels are inherently unstable, with computer control they perform stable three and four wheel setups. How many more “active” devices are there yet to be discovered?

More information

Climb

Design and control of an in-flight reconfigurable quadcopter using non-actuated hinges

by Nathan Bucki, Jerry Tang and Mark W. Mueller.

Related Articles

Google adds Boston Dynamics to its robotics acquisitions

Boston Dynamics & Shaft sold to SoftBank

Do you love me? Boston Dynamics robots show how good they are

Atlas the Gymnast, Spot the Quadruped

To be informed of new articles on I Programmer, subscribe to our weekly newsletter, subscribe to the RSS feed and follow us on Twitter, Facebook Where LinkedIn.

Banner

square

comments

or send your comment to: [email protected]ogrammer.info

Share.

Comments are closed.