Tag Archives: robot
Researchers at University of Washington have recently developed a new protocol to train robots and test their performance on tasks that involve object manipulation. This protocol, presented in a paper published in IEEE Robotics and Automation Letters, is based on the Rubik's Cube, the well-known 3D combination puzzle invented by the Hungarian sculpture and architect Ernő Rubik. Continue reading
Video Friday is your weekly selection of awesome robotics videos, collected by your friends at IEEE Spectrum robotics. We’ll also be posting a weekly calendar of upcoming robotics events for the next few months; here’s what we have so far (send us your events!):
ICRA 2022: 23–27 May 2022, PhiladelphiaERF 2022: 28–30 June 2022, Rotterdam, Netherlands CLAWAR 2022: 12–14 September 2022, Azores, PortugalLet us know if you have suggestions for next week, and enjoy today’s videos.
Kawasaki introduced a bunch of new robots at iREX in Japan. Including this friendly fellow, called Kaleido.
And here, of course, is Bex, the robotic ibex that we wrote about earlier this week. But it’s way weirder on video!
川崎重工 「RHP Bex」人が乗る #2022国際ロボット展 #irex2022
There are some other new robots too! Robotstart was there in person, so check out its coverage at the link below.
[ Robotstart ]
In better times, Ukrainian drone enthusiasts flew their gadgets into the sky to photograph weddings or race drones for fun. Now some are risking their lives by forming a volunteer drone force to help their country repel the Russian invasion.
[ AP News ]
Now on Kickstarter, a robotic cat that will not make you sneeze!
As always, remember that crowdfunding videos like these show only the best possible performance and demonstrate a borderline unrealistic experience. Kickstart accordingly, starting at $1K.
[ Kickstarter ]
This is a pretty great application for a robot arm, right?
The obvious next step is making the driving autonomous too, so that the operator can take a nap.
[ RMV ]
DRC-Hubo has joined the UMass Lowell NERVE Center, where it’s going to learn how to do useful stuff on boats. I totally get why you’d want (or need!) a biped for boatvironments, but good luck with balancing on the high seas.
[ UML ]
Marsupial robot teams have lots of advantages when exploring unknown environments, as long as you can keep them from stepping on or falling off each other. I love the bungie idea here.
[ ARL ]
Fly Like A Girl is a half-day program that introduces youth aged 10–18 to different aspects of drones and applications behind them.
[ WeRobotics ]
iCub embracing its well-documented obsession with condiments.
[ Paper ]
Quadrotors, meet robot arm!
[ INRoL ]
UAVs have a lot to learn from birds. Although, birds may be able to learn a thing or two from UAVs as well, like how to sound like a swarm of giant bees.
[ UMich ]
Haptic teleoperation of a 6 DoF omnidirectional drone sure seems like a very challenging controls problem.
[ ASL ]
Did you ever hide something in the sand at the beach when you were a kid? This is like that, except for adults and robots.
[ DFKI ]
A mobile robot that can explore, inspect, and apply air-sealing foam in spaces beyond human reach.
[ Northeastern ]
Telexistence, best known for its futuristic telepresence anime humanoid robots, is now working on something decidely more mundane.
[ Telexistence ]
In this work, we make two contributions: (i) we perform the first benchmark comparison of existing learned control policies for agile quadrotor flight and show that training a control policy that commands body-rates and thrust results in more robust sim-to-real transfer compared to a policy that directly specifies individual rotor thrusts, (ii) we demonstrate for the first time that such a control policy trained via deep reinforcement learning can control a quadrotor in real-world experiments at speeds over 45km/h.
[ UZH RPG ]
Carmel Majidi, Sarah Bergbreiter, and Vickie Webster-Wood talk about the vision for Softbotics: to engineer machines and robots that put humans at the center and are designed for everyday life.
[ CMU ] Continue reading
This is Bex. It is, as near as I can tell, a robotic ibex, designed by Kawasaki and inspired by the fiercely horny species of goat native to many mountainous regions of Africa and Eurasia. Bex made its first appearance at iREX this week in Japan, and we should talk about why a Kawasaki robot ibex now exists, besides, of course, the simple fact that it's just so darn fun to say out loud.
Kawasaki has been working on a “Robust Humanoid Platform” (RHP) called Kaleido since 2015, and Bex is a “friend” of that program. Masayuki Soube, who is in charge of development of the RHP, was interviewed recently on Kawasaki Robotics' website about the Bex program. The conversation was conducted in Japanese. What follows (in italics) are Google Translate's best attempts at rendering Soube's observations into workable English:
Through the development of Kaleido, we felt the difficulty of biped robots. Because humanoid robots have the same shape as humans, they are highly versatile, with the potential to do everything that humans can ultimately do. However, it will take a long time to put it to practical use. On the other hand, we are also developing a self-propelled service robot that moves on wheels, but legs are still suitable for moving on rough terrain rather than wheels. So, halfway between humanoid robots and wheeled robots, [we]l wondered if there was an opportunity. That's why we started developing Bex, a quadruped walking robot. We believe that the walking technology cultivated in the development of humanoid robots can definitely be applied to quadruped walking robots.
It's worth noting that Bex appears to have a hybrid mobility system, with knee wheels that it can kneel on to move more quickly on smooth and level surfaces. It's also worth noting that Bex has a seat and handlebars (!), and it can carry up to 100 kilograms.
Rather than committing Bex to human hauling, Kawasaki had sadly settled on more practical applications, according to Masayuki Soube:
First of all Bex can carry light cargo, such as transporting materials at a construction site. The other application is inspection. In a vast industrial plant, Bex can look around and images from its camera can be remotely checked to see what the instruments are doing. Bex can also carry crops harvested by humans on farmland.
Although the base of Bex is a legged robot, the upper body of Bex is not fixed and we are thinking of adapting it according to the application. If it is a construction site, we will form a partnership with the construction manufacturer, and if it is a plant, we will form a partnership with the plant manufacturer and leave the upper body to us. Kawasaki Heavy Industries will focus on the four legs of the lower body and want to provide it as an open innovation platform.
Bex under development at Kawasaki RoboticsKawasaki Robotics
Kawasaki is certainly not the first company to be thinking about quadrupedal robots as inspection platforms. They're also not the first to work on hybrid mobility for quadrupeds; in fact, I'm betting that Bex would have a lot of (literal) catching up to do in that space. But the idea of a quadrupedal cargo system is somewhat interesting—most quadrupeds consider cargo in terms of what useful payload they can carry, and typically those payloads are engineered around things like sensors, which don't weigh all that much. Could Bex bring us back to the days of Boston Dynamics' BigDog and LS3, where quadrupedal cargo haulers took a run at practicality? It would take a lot of work, but boy do I hope so. Big, powerful quadrupeds are super cool, especially if they have horns. Continue reading
Over the past decades, engineers have created devices with increasingly advanced functions and capabilities. A device capability that was substantially improved in recent years is known as “spatial computing.” Continue reading
A team of researchers at University of California, Los Angeles (UCLA)'s Center for Vision, Cognition, Learning, and Autonomy (VCLA), led by Prof. Song-Chun Zhu, recently developed an approach that could help to align a human user's assessment of what a robot can do with its true capabilities. This approach, presented in a paper published in IEEE Robotics and Automation Letters, is based on a new algorithm that simultaneously optimizes the physical cost and expressiveness of a robot's motion, to determine how well human observers would estimate its reachable workspace. Continue reading