Your weekly selection of awesome robot videos
Video Friday is your weekly selection of awesome robotics videos, collected by your Automaton bloggers. 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!):
IEEE IRC 2018 – January 31-2, 2018 – Laguna Hills, Calif.
International Symposium on Medical Robotics – March 1-3, 2018 – Atlanta, Georgia
HRI 2018 – March 5-8, 2018 – Chicago, Ill.
Let us know if you have suggestions for next week, and enjoy today’s videos.
CES isn’t really a venue for the launch of flagship robotics products anymore (if it ever was), but we still see some high profile introductions from large companies looking to make a splash. Besides LG, Honda was the other notable, with a couple strange robots (and one kind of familiar looking). Here’s an intro:
Now that you’ve seen the robots themselves, here’s a weird concept video showing Honda’s vision for these things:
The most interesting, I think, is the 3E-A18, so here’s a bit more on that one:
And the 3E-D18? That looks rather a lot like Clearpath’s old Grizzly UGV:
[ Honda ]
The Keplinger Research Group at the University of Colorado Boulder has developed a new class of soft electrically activated artificial muscles, called HASEL (Hydraulically Amplified Self-healing ELectrostatic) actuators, which exceed or match the strength, speed, and efficiency of biological muscle. HASEL devices are powerful enough to lift a gallon of water, yet offer the versatility to perform delicate tasks like grasping small fruit or an egg. Also, HASEL can self-heal from electrical damage, self-sense position, and operate at high-speeds, all while being made from inexpensive materials like silicone rubber, hydrogel, and vegetable oil.
In less than three months, a team of engineers designed and built the unmanned cargo air vehicle (CAV) prototype. The innovative vehicle will be used to test and evolve Boeing’s autonomous technology for future aerospace vehicles.
Powered by an environmentally-friendly electric propulsion system, the CAV prototype is outfitted with eight counter rotating blades allowing for vertical flight. It measures 15 feet long (4.57 meters), 18 feet wide (5.49 meters) and 4 feet tall (1.22 meters), and weighs 747 pounds (339 kilograms). It is designed to transport a payload up to 500 pounds for possible future cargo and logistics applications.
Honestly, this is the most compelling delivery drone prototype I’ve seen in a long time.
[ Boeing ]
I do not understand the appeal of things like this.
I mean, even the CGI version doesn’t really look that exciting? The problem with these big robots with humans inside them is that a.) the humans are basically redundant and b.) robots that size don’t move in dynamic or interesting ways. They just lumber slowly and break a lot. For the sake of argument, let’s say that these things do end up moving at 20 mph—great, then what happens…? I guess maybe if they threw in a bunch of obstacles where the legs would be able to do unique leg stuff, then okay, but otherwise, to me this just comes off as something you should probably have put wheels or tracks on instead.
[ Furrion ]
The Torc self-driving car, Asimov, performs a U-turn and encounters a car driving the wrong direction in its path — at the same time pedestrians are in the road and another car overlaps the lane. The Torc car slows to a stop in front of a parking lot entrance, allowing the wrong-way vehicle to turn off the road.
It’s usually a safe thing for a self driving car to just slow down and stop when it encounters an edge case like this. However, as a human driver, I probably would have stopped much earlier and not crept forward like Asimov did. Incidentally, I wonder if the car can tell the difference between the front of a vehicle and the back of a vehicle? Like, did it know that this was a car facing the wrong way and going forwards, rather than a car facing the right way and going backwards?
[ TORC Robotics ]
Misty Robotics held a robothon with Misty I prototypes back in December, and none of y’all snuck me a picture >:(
[ Misty Robotics ]
The intelligent security robot is Nimbo from Turing Video Inc., powered by Segway Robotics. With industry-leading sensors on board, such as Intel RealSense, the Segway Robotics base provides stability and flexibility for indoor and outdoor use. It can handle rough or uneven pavement and move over speed bumps, while the relatively small size (30-inch height and 25-inch width) enables it to fit in narrow spaces. Additionally, the robot can be switched to the “Ride-On Mode”, a distinctive feature that allows a security staff member to step on and ride it up to 11 mph.
The ROVéo Fire is an interesing design for a unmanned ground vehicle that relies on mechanical design to surmount obstacles:
[ Rovenso ]
We propose a novel method to deal with the scale ambiguity in monocular SLAM based on control stability. We analytically show that (1) using unscaled state feedback from monocular SLAM for control can lead to system instability, and (2) there is a unique linear relationship between the absolute scale of the SLAM system and the control gain at which such instability arises. Using this property, our system estimates the scale by adapting the gain and detecting the self-induced oscillations. Unlike conventional monocular approaches, no additional metric sensors are used for scale estimation. We demonstrate the ability of our method to estimate the scale for performing autonomous indoor navigation with a low-cost quadrotor MAV.
Work from the MavLab at TU Delft; paper at the link below.
Dhanvin Mehta, CSE PhD Student, is working on the Risk Aware Multi-Policy Decision Making algorithm to help guide robots in uncertain environments with constantly changing variables, such as for autonomous vehicles driving among pedestrians who can suddenly change direction or speed without signalling. The algorithm takes into account potentially dangerous outcomes in the near future and adjusts to each unique situation, making the best decision possible.
[ APRIL ]
Howie Choset’s Robot Kinematics course at CMU is teaching a robot arm to stack Jenga blocks. Sort of. The actual stacking starts at 1:40ish.
[ CMU ]
Want to work at iRobot? Seems like it would be a pretty good gig:
Although, you know, Boston.
[ iRobot ]
Advances in hardware and software are happening as quickly as ever in robotics, and the RoboSub teams are benefiting just as much as anyone else:
[ RoboSub ]
In the latest episode of Robots in Depth, Per interviews Erin Rapacki, founder of Machine Inbound.
Erin talks to Per about her work in marketing for several startups. They discuss selling points, including examples from different industries, and marketing for launching a robotics product. Erin also shares her insights on telepresence robots from working with Beam and her game plan for starting a new company based on robotics technology. We get Erin’s perspective on IOT and robotics being the next wave of startups and that venture capital is adapting to the difference between web and app investments on one hand and hardware and IOT/robotics on the other. Erin thinks that robotics will be introduced in specific verticals and that the service industry will be early adopters. We also find out about how the robot loving customer is her biggest problem!
[ Robots in Depth ]
Source: IEEE Spectrum