Close

Video Monday: Agile Justin, Baby Elephant Robot, and More From ICRA 2014

While we put together more ICRA posts, here are some of the best videos for you

4 min read
Video Monday: Agile Justin, Baby Elephant Robot, and More From ICRA 2014

We're back from the 2014 IEEE International Conference on Robotics and Automation (ICRA) in Hong Kong, and as always, there was a stupendous amount of incredible research that was presented across three days of conference and two days of workshops. We've already posted a bunch of cool stuff, and we've got more in the works, but for now, here's a stack of awesome research videos for you to have a look at.

 

 

This video of DLR's Agile Justin won the Best Video Award at ICRA. It shows off some of the hardware and capabilities of what's arguably one of the most, if not the most, capable dual-armed mobile humanoid robots in existence.

This video presents the recent upgrades of DLR's mobile humanoid Agile Justin, bringing it closer to an ideal platform for research in autonomous manipulation. Significant upgrades have been made in the fields of mechatronics, 3D sensors, tactile skin, massive GPGPU based computing power, and software communication framework. In addition, first algorithms and two experimental scenarios are presented that take advantage of these new capabilities.

"Agile Justin: An Upgraded Member of DLR’s Family of Lightweight and Torque Controlled Humanoids," by B. Bauml, T. Hammer, R. Wagner, O. Birbach, T. Gumpert, F. Zhi, U. Hillenbrand, S. Beer, W. Friedl, and J. Butterfass, from the DLR Institute of Robotics and Mechatronics in Germany, was presented at ICRA 2014 in Hong Kong.

 

 

The Kuka Best Service Robotics Paper Award went to a paper describing a dual-armed robot able to robustly unfold a variety of clothing. As we've seen, it's not folding clothes that's the challenge, it's first unfolding them from the jumbled pile that comes out of a dryer. The robot relies only on gravity, vision, and a minimal number of grasps to recognize and unfold long sleeve shirts, short sleeve shirts, pants, and shorts. The work is a collaboration between the Information Technologies Institute (ITI) in Greece and Imperial College London.

"Autonomous Active Recognition and Unfolding of Clothes using Random Decision Forests and Probabilistic Planning," by Andreas Doumanoglou, Andreas Kargakos, Tae-Kyun Kim, and Sotiris Malassiotis from ITI and Imperial College London, was presented at ICRA 2014 in Hong Kong. 

 

 

Turboquad is a robot that can give you wheels when you want them, or legs (of a sort) when wheels won't do. It's an evolution of the Quattroped robot that we covered several years ago:

"TurboQuad: A Leg-Wheel Transformable Robot Using Bio-Inspired Control," by Wei-Hsi Chen, Hung-Sheng Lin, and Pei-Chun Lin, from National Taiwan University, was presented at ICRA 2014 in Hong Kong.

 

 

Another follow-up to research that we've covered in the past is this cheetah-inspired wheeled robot, which used an actuated tail to allow it to take very sharp turns at high speed. At ICRA, the researchers presented a variation that uses a tail to help control rapid acceleration and braking:

Stimulated by observations of cheetahs accelerating, a novel tail controller system for rapid acceleration and braking is presented. To understand the targeted behaviour of a cheetah performing a longitudinal manoeuvre and the effects an actuated tail has, a simple mathematical template was developed. Subsequently feedback controllers were designed based on our hypothesis. Finally, the control system was experimentally tested on a reduced order robot model which increased its manoeuvrability considerably.

"Rapid Acceleration and Braking: Inspirations from the Cheetah’s Tail," by Amir Patel and M. Braae from the University of Cape Town, South Africa, was presented at ICRA 2014 in Hong Kong.

 

 

Robots are complicated things, with motors and servos and stuff all over the place. This makes them somewhat sensitive to excessive amounts of water, dust, and mud. One way to get around this problem is to build a robot that is completely sealed, and we mean completely sealed: this robot, called Torus, is covered in a seamless skin that it drives inside to move. Stick around to the end of the video to see if driving wire-free.

"Torus Omnidirectional Driving Unit Mechanism Realized by Curved Crawler Belts," by Kenjiro Tadakuma, Hirohiko Ogata, Riichiro Tadakuma and Jose Berengueres from Osaka University, was presented at ICRA 2014 in Hong Kong.

 

 

Legged robots can adapt to different terrain or travel at different speeds by choosing to use different gaits. Underwater robots can also take advantage of different gaits, if you give them limbs and enough degrees of freedom. This is Glide Walker:

"Gliding, Swimming and Walking: Development of multi-functional underwater robot Glide Walker," by Hirotaka Komura, Satoshi Kitano, Hiroya Yamada, and Gen Endo from Tokyo Institute of Technology, was presented at ICRA 2014 in Hong Kong.

 

 

When we (humans) pick something up, we may not always get our grasp correct on the first try. Or, we may need to change our grasp for some reason, if we're using a tool, for example. This is a skill that robots are going to have to master if they're going to operate in human environments, and ultimately, they may develop the skills to become better at it than humans are. Researchers from MIT, Carnegie Mellon, and ABB are exploring a concept called "extrinsic dexterity," using passive and active techniques to regrasp objects:

Dexterity is not a property of a robot hand, but that of the entire system. This video demonstrates that dexterous manipulation is possible with a hand dramatically simpler than typical dexterous hands. The key is to use the motions of the arm, object inertia, gravity, and external contacts: extrinsic dexterity.

The video showcases a repertoire of regrasps developed for a simple gripper (MLab Hand) and presents one of the sequences of regrasps designed to explore border manipulation capability by connecting different regrasps.

"Regrasping Objects Using Extrinsic Dexterity," by Nikhil Chavan Dafle, Alberto Rodriguez, Robert Paolini, Bowei Tang, Siddhartha S. Srinivasa, Michael Erdmann, Matthew T. Mason, Ivan Lundberg, Harald Staab, and Thomas Fuhlbrigge from MIT, Carnegie Mellon, and ABB, was presented at ICRA 2014 in Hong Kong.

 

 

Finally! A walking robot designed for people to ride! Check out the saddle on this "Baby Elephant" at 2:30 in:

This video introduces our lately developed quadruped robot. It is named as “Baby Elephant” because of the heavy load capability and the elephant like appearance. The leg is a serial-parallel hybrid mechanism that is one novelty of our robot. The maximum speed is 1.8 km/h. The static gait experiment with the same load was also conducted, which shows that the Baby Elephant can walk on different kinds of terrains. The maximum load can be up to 100kg. 

"A quadruped robot with parallel mechanism legs," by Feng Gao, Chenkun Qi, Qiao Sun, Xianbao Chen and Xinghua Tian from Shanghai Jiao Tong University, was presented at ICRA 2014 in Hong Kong.

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman
LightGreen

This article is part of our special report on AI, “The Great AI Reckoning.

"I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓ Show less