Tag Archives: short
#437639 Boston Dynamics’ Spot Is Helping ...
In terms of places where you absolutely want a robot to go instead of you, what remains of the utterly destroyed Chernobyl Reactor 4 should be very near the top of your list. The reactor, which suffered a catastrophic meltdown in 1986, has been covered up in almost every way possible in an effort to keep its nuclear core contained. But eventually, that nuclear material is going to have to be dealt with somehow, and in order to do that, it’s important to understand which bits of it are just really bad, and which bits are the actual worst. And this is where Spot is stepping in to help.
The big open space that Spot is walking through is right next to what’s left of Reactor 4. Within six months of the disaster, Reactor 4 was covered in a sarcophagus made of concrete and steel to try and keep all the nasty nuclear fuel from leaking out more than it already had, and it still contains “30 tons of highly contaminated dust, 16 tons of uranium and plutonium, and 200 tons of radioactive lava.” Oof. Over the next 10 years, the sarcophagus slowly deteriorated, and despite the addition of that gigantic network of steel support beams that you can see in the video, in the late 1990s it was decided to erect an enormous building over the entire mess to try and stabilize it for as long as possible.
Reactor 4 is now snugly inside the massive New Safe Confinement (NSC) structure, and the idea is that eventually, the structure will allow for the safe disassembly of what’s left of the reactor, although nobody is quite sure how to do that. This is all just to say that the area inside of the containment structure offers a lot of good opportunities for robots to take over from humans.
This particular Spot is owned by the U.K. Atomic Energy Authority, and was packed off to Russia with the assistance of the Robotics and Artificial Intelligence in Nuclear (RAIN) initiative and the National Centre for Nuclear Robotics. Dr. Dave Megson-Smith, who is a researcher at the University of Bristol, in the U.K., and part of the Hot Robotics Facility at the National Nuclear User Facility, was one of the scientists lucky enough to accompany Spot on its adventure. Megson-Smith specializes in sensor development, and he equipped Spot with a collimated radiation sensor in addition to its mapping payload. “We actually built a map of the radiation coming out of the front wall of Chernobyl power plant as we were in there with it,” Megson-Smith told us, and was able to share this picture, which shows a map of gamma photon count rate:
Image: University of Bristol
Researchers equipped Spot with a collimated radiation sensor and use one of the data readings (gamma photon count rate) to create a map of the radiation coming out of the front wall of the Chernobyl power plant.
So what’s the reason you’d want to use a very expensive legged robot to wander around what looks like a very flat and robot friendly floor? As it turns out, the floor is very dusty in there, and a priority inside the NSC is to keep dust down as much as possible, since the dust is radioactive and gets on everything and is consequently the easiest way for radioactivity to escape the NSC. “You want to minimize picking up material, so we consider the total contact surface area,” says Megson-Smith. “If you use a legged system rather than a wheeled or tracked system, you have a much smaller footprint and you disturb the environment a lot less.” While it’s nice that Spot is nimble and can climb stairs and stuff, tracked vehicles can do that as well, so in this case, the primary driving factor of choosing a robot to work inside Chernobyl is minimizing those contact points.
Right now, routine weekly measurements in contaminated spaces at Chernobyl are done by humans, which puts those humans at risk. Spot, or a robot like it, could potentially take over from those humans, as a sort of “automated safety checker”
Right now, routine weekly measurements in contaminated spaces at Chernobyl are done by humans, which puts those humans at risk. Spot, or a robot like it, could potentially take over from those humans, as a sort of “automated safety checker” able to work in medium level contaminated environments.” As far as more dangerous areas go, there’s a lot of uncertainty about what Spot is actually capable of, according to Megson-Smith. “What you think the problems are, and what the industry thinks the problems are, are subtly different things.
We were thinking that we’d have to make robots incredibly radiation proof to go into these contaminated environments, but they said, “can you just give us a system that we can send into places where humans already can go, but where we just don’t want to send humans.” Making robots incredibly radiation proof is challenging, and without extensive testing and ruggedizing, failures can be frequent, as many robots discovered at Fukushima. Indeed, Megson-Smith that in Fukushima there’s a particular section that’s known as a “robot graveyard” where robots just go to die, and they’ve had to up their standards again and again to keep the robots from failing. “So the thing they’re worried about with Spot is, what is its tolerance? What components will fail, and what can we do to harden it?” he says. “We’re approaching Boston Dynamics at the moment to see if they’ll work with us to address some of those questions.
There’s been a small amount of testing of how robots fair under harsh radiation, Megson-Smith told us, including (relatively recently) a KUKA LBR800 arm, which “stopped operating after a large radiation dose of 164.55(±1.09) Gy to its end effector, and the component causing the failure was an optical encoder.” And in case you’re wondering how much radiation that is, a 1 to 2 Gy dose to the entire body gets you acute radiation sickness and possibly death, while 8 Gy is usually just straight-up death. The goal here is not to kill robots (I mean, it sort of is), but as Megson-Smith says, “if we can work out what the weak points are in a robotic system, can we address those, can we redesign those, or at least understand when they might start to fail?” Now all he has to do is convince Boston Dynamics to send them a Spot that they can zap until it keels over.
The goal for Spot in the short term is fully autonomous radiation mapping, which seems very possible. It’ll also get tested with a wider range of sensor packages, and (happily for the robot) this will all take place safely back at home in the U.K. As far as Chernobyl is concerned, robots will likely have a substantial role to play in the near future. “Ultimately, Chernobyl has to be taken apart and decommissioned. That’s the long-term plan for the facility. To do that, you first need to understand everything, which is where we come in with our sensor systems and robotic platforms,” Megson-Smith tells us. “Since there are entire swathes of the Chernobyl nuclear plant where people can’t go in, we’d need robots like Spot to do those environmental characterizations.” Continue reading →
#437624 AI-Powered Drone Learns Extreme ...
Quadrotors are among the most agile and dynamic machines ever created. In the hands of a skilled human pilot, they can do some astonishing series of maneuvers. And while autonomous flying robots have been getting better at flying dynamically in real-world environments, they still haven’t demonstrated the same level of agility of manually piloted ones.
Now researchers from the Robotics and Perception Group at the University of Zurich and ETH Zurich, in collaboration with Intel, have developed a neural network training method that “enables an autonomous quadrotor to fly extreme acrobatic maneuvers with only onboard sensing and computation.” Extreme.
There are two notable things here: First, the quadrotor can do these extreme acrobatics outdoors without any kind of external camera or motion-tracking system to help it out (all sensing and computing is onboard). Second, all of the AI training is done in simulation, without the need for an additional simulation-to-real-world (what researchers call “sim-to-real”) transfer step. Usually, a sim-to-real transfer step means putting your quadrotor into one of those aforementioned external tracking systems, so that it doesn’t completely bork itself while trying to reconcile the differences between the simulated world and the real world, where, as the researchers wrote in a paper describing their system, “even tiny mistakes can result in catastrophic outcomes.”
To enable “zero-shot” sim-to-real transfer, the neural net training in simulation uses an expert controller that knows exactly what’s going on to teach a “student controller” that has much less perfect knowledge. That is, the simulated sensory input that the student ends up using as it learns to follow the expert has been abstracted to present the kind of imperfect, imprecise data it’s going to encounter in the real world. This can involve things like abstracting away the image part of the simulation until you’d have no way of telling the difference between abstracted simulation and abstracted reality, which is what allows the system to make that sim-to-real leap.
The simulation environment that the researchers used was Gazebo, slightly modified to better simulate quadrotor physics. Meanwhile, over in reality, a custom 1.5-kilogram quadrotor with a 4:1 thrust to weight ratio performed the physical experiments, using only a Nvidia Jetson TX2 computing board and an Intel RealSense T265, a dual fisheye camera module optimized for V-SLAM. To challenge the learning system, it was trained to perform three acrobatic maneuvers plus a combo of all of them:
Image: University of Zurich/ETH Zurich/Intel
Reference trajectories for acrobatic maneuvers. Top row, from left: Power Loop, Barrel Roll, and Matty Flip. Bottom row: Combo.
All of these maneuvers require high accelerations of up to 3 g’s and careful control, and the Matty Flip is particularly challenging, at least for humans, because the whole thing is done while the drone is flying backwards. Still, after just a few hours of training in simulation, the drone was totally real-world competent at these tricks, and could even extrapolate a little bit to perform maneuvers that it was not explicitly trained on, like doing multiple loops in a row. Where humans still have the advantage over drones is (as you might expect since we’re talking about robots) is quickly reacting to novel or unexpected situations. And when you’re doing this sort of thing outdoors, novel and unexpected situations are everywhere, from a gust of wind to a jealous bird.
For more details, we spoke with Antonio Loquercio from the University of Zurich’s Robotics and Perception Group.
IEEE Spectrum: Can you explain how the abstraction layer interfaces with the simulated sensors to enable effective sim-to-real transfer?
Antonio Loquercio: The abstraction layer applies a specific function to the raw sensor information. Exactly the same function is applied to the real and simulated sensors. The result of the function, which is “abstracted sensor measurements,” makes simulated and real observation of the same scene similar. For example, suppose we have a sequence of simulated and real images. We can very easily tell apart the real from the simulated ones given the difference in rendering. But if we apply the abstraction function of “feature tracks,” which are point correspondences in time, it becomes very difficult to tell which are the simulated and real feature tracks, since point correspondences are independent of the rendering. This applies for humans as well as for neural networks: Training policies on raw images gives low sim-to-real transfer (since images are too different between domains), while training on the abstracted images has high transfer abilities.
How useful is visual input from a camera like the Intel RealSense T265 for state estimation during such aggressive maneuvers? Would using an event camera substantially improve state estimation?
Our end-to-end controller does not require a state estimation module. It shares however some components with traditional state estimation pipelines, specifically the feature extractor and the inertial measurement unit (IMU) pre-processing and integration function. The input of the neural networks are feature tracks and integrated IMU measurements. When looking at images with low features (for example when the camera points to the sky), the neural net will mainly rely on IMU. When more features are available, the network uses to correct the accumulated drift from IMU. Overall, we noticed that for very short maneuvers IMU measurements were sufficient for the task. However, for longer ones, visual information was necessary to successfully address the IMU drift and complete the maneuver. Indeed, visual information reduces the odds of a crash by up to 30 percent in the longest maneuvers. We definitely think that event camera can improve even more the current approach since they could provide valuable visual information during high speed.
“The Matty Flip is probably one of the maneuvers that our approach can do very well … It is super challenging for humans, since they don’t see where they’re going and have problems in estimating their speed. For our approach the maneuver is no problem at all, since we can estimate forward velocities as well as backward velocities.”
—Antonio Loquercio, University of Zurich
You describe being able to train on “maneuvers that stretch the abilities of even expert human pilots.” What are some examples of acrobatics that your drones might be able to do that most human pilots would not be capable of?
The Matty Flip is probably one of the maneuvers that our approach can do very well, but human pilots find very challenging. It basically entails doing a high speed power loop by always looking backward. It is super challenging for humans, since they don’t see where they’re going and have problems in estimating their speed. For our approach the maneuver is no problem at all, since we can estimate forward velocities as well as backward velocities.
What are the limits to the performance of this system?
At the moment the main limitation is the maneuver duration. We never trained a controller that could perform maneuvers longer than 20 seconds. In the future, we plan to address this limitation and train general controllers which can fly in that agile way for significantly longer with relatively small drift. In this way, we could start being competitive against human pilots in drone racing competitions.
Can you talk about how the techniques developed here could be applied beyond drone acrobatics?
The current approach allows us to do acrobatics and agile flight in free space. We are now working to perform agile flight in cluttered environments, which requires a higher degree of understanding of the surrounding with respect to this project. Drone acrobatics is of course only an example application. We selected it because it makes a stress test of the controller performance. However, several other applications which require fast and agile flight can benefit from our approach. Examples are delivery (we want our Amazon packets always faster, don’t we?), search and rescue, or inspection. Going faster allows us to cover more space in less time, saving battery costs. Indeed, agile flight has very similar battery consumption of slow hovering for an autonomous drone.
“Deep Drone Acrobatics,” by Elia Kaufmann, Antonio Loquercio, René Ranftl, Matthias Müller, Vladlen Koltun, and Davide Scaramuzza from the Robotics and Perception Group at the University of Zurich and ETH Zurich, and Intel’s Intelligent Systems Lab, was presented at RSS 2020. Continue reading →
#437620 The Trillion-Transistor Chip That Just ...
The history of computer chips is a thrilling tale of extreme miniaturization.
The smaller, the better is a trend that’s given birth to the digital world as we know it. So, why on earth would you want to reverse course and make chips a lot bigger? Well, while there’s no particularly good reason to have a chip the size of an iPad in an iPad, such a chip may prove to be genius for more specific uses, like artificial intelligence or simulations of the physical world.
At least, that’s what Cerebras, the maker of the biggest computer chip in the world, is hoping.
The Cerebras Wafer-Scale Engine is massive any way you slice it. The chip is 8.5 inches to a side and houses 1.2 trillion transistors. The next biggest chip, NVIDIA’s A100 GPU, measures an inch to a side and has a mere 54 billion transistors. The former is new, largely untested and, so far, one-of-a-kind. The latter is well-loved, mass-produced, and has taken over the world of AI and supercomputing in the last decade.
So can Goliath flip the script on David? Cerebras is on a mission to find out.
Big Chips Beyond AI
When Cerebras first came out of stealth last year, the company said it could significantly speed up the training of deep learning models.
Since then, the WSE has made its way into a handful of supercomputing labs, where the company’s customers are putting it through its paces. One of those labs, the National Energy Technology Laboratory, is looking to see what it can do beyond AI.
So, in a recent trial, researchers pitted the chip—which is housed in an all-in-one system about the size of a dorm room mini-fridge called the CS-1—against a supercomputer in a fluid dynamics simulation. Simulating the movement of fluids is a common supercomputer application useful for solving complex problems like weather forecasting and airplane wing design.
The trial was described in a preprint paper written by a team led by Cerebras’s Michael James and NETL’s Dirk Van Essendelft and presented at the supercomputing conference SC20 this week. The team said the CS-1 completed a simulation of combustion in a power plant roughly 200 times faster than it took the Joule 2.0 supercomputer to do a similar task.
The CS-1 was actually faster-than-real-time. As Cerebrus wrote in a blog post, “It can tell you what is going to happen in the future faster than the laws of physics produce the same result.”
The researchers said the CS-1’s performance couldn’t be matched by any number of CPUs and GPUs. And CEO and cofounder Andrew Feldman told VentureBeat that would be true “no matter how large the supercomputer is.” At a point, scaling a supercomputer like Joule no longer produces better results in this kind of problem. That’s why Joule’s simulation speed peaked at 16,384 cores, a fraction of its total 86,400 cores.
A comparison of the two machines drives the point home. Joule is the 81st fastest supercomputer in the world, takes up dozens of server racks, consumes up to 450 kilowatts of power, and required tens of millions of dollars to build. The CS-1, by comparison, fits in a third of a server rack, consumes 20 kilowatts of power, and sells for a few million dollars.
While the task is niche (but useful) and the problem well-suited to the CS-1, it’s still a pretty stunning result. So how’d they pull it off? It’s all in the design.
Cut the Commute
Computer chips begin life on a big piece of silicon called a wafer. Multiple chips are etched onto the same wafer and then the wafer is cut into individual chips. While the WSE is also etched onto a silicon wafer, the wafer is left intact as a single, operating unit. This wafer-scale chip contains almost 400,000 processing cores. Each core is connected to its own dedicated memory and its four neighboring cores.
Putting that many cores on a single chip and giving them their own memory is why the WSE is bigger; it’s also why, in this case, it’s better.
Most large-scale computing tasks depend on massively parallel processing. Researchers distribute the task among hundreds or thousands of chips. The chips need to work in concert, so they’re in constant communication, shuttling information back and forth. A similar process takes place within each chip, as information moves between processor cores, which are doing the calculations, and shared memory to store the results.
It’s a little like an old-timey company that does all its business on paper.
The company uses couriers to send and collect documents from other branches and archives across town. The couriers know the best routes through the city, but the trips take some minimum amount of time determined by the distance between the branches and archives, the courier’s top speed, and how many other couriers are on the road. In short, distance and traffic slow things down.
Now, imagine the company builds a brand new gleaming skyscraper. Every branch is moved into the new building and every worker gets a small filing cabinet in their office to store documents. Now any document they need can be stored and retrieved in the time it takes to step across the office or down the hall to their neighbor’s office. The information commute has all but disappeared. Everything’s in the same house.
Cerebras’s megachip is a bit like that skyscraper. The way it shuttles information—aided further by its specially tailored compiling software—is far more efficient compared to a traditional supercomputer that needs to network a ton of traditional chips.
Simulating the World as It Unfolds
It’s worth noting the chip can only handle problems small enough to fit on the wafer. But such problems may have quite practical applications because of the machine’s ability to do high-fidelity simulation in real-time. The authors note, for example, the machine should in theory be able to accurately simulate the air flow around a helicopter trying to land on a flight deck and semi-automate the process—something not possible with traditional chips.
Another opportunity, they note, would be to use a simulation as input to train a neural network also residing on the chip. In an intriguing and related example, a Caltech machine learning technique recently proved to be 1,000 times faster at solving the same kind of partial differential equations at play here to simulate fluid dynamics.
They also note that improvements in the chip (and others like it, should they arrive) will push back the limits of what can be accomplished. Already, Cerebras has teased the release of its next-generation chip, which will have 2.6 trillion transistors, 850,00 cores, and more than double the memory.
Of course, it still remains to be seen whether wafer-scale computing really takes off. The idea has been around for decades, but Cerebras is the first to pursue it seriously. Clearly, they believe they’ve solved the problem in a way that’s useful and economical.
Other new architectures are also being pursued in the lab. Memristor-based neuromorphic chips, for example, mimic the brain by putting processing and memory into individual transistor-like components. And of course, quantum computers are in a separate lane, but tackle similar problems.
It could be that one of these technologies eventually rises to rule them all. Or, and this seems just as likely, computing may splinter into a bizarre quilt of radical chips, all stitched together to make the most of each depending on the situation.
Image credit: Cerebras Continue reading →
#437603 Throwable Robot Car Always Lands on Four ...
Throwable or droppable robots seem like a great idea for a bunch of applications, including exploration and search and rescue. But such robots do come with some constraints—namely, if you’re going to throw or drop a robot, you should be prepared for that robot to not land the way you want it to land. While we’ve seen some creative approaches to this problem, or more straightforward self-righting devices, usually you’re in for significant trade-offs in complexity, mobility, and mass.
What would be ideal is a robot that can be relied upon to just always land the right way up. A robotic cat, of sorts. And while we’ve seen this with a tail, for wheeled vehicles, it turns out that a tail isn’t necessary: All it takes is some wheel spin.
The reason that AGRO (Agile Ground RObot), developed at the U.S. Military Academy at West Point, can do this is because each of its wheels is both independently driven and steerable. The wheels are essentially reaction wheels, which are a pretty common way to generate forces on all kinds of different robots, but typically you see such reaction wheels kludged onto these robots as sort of an afterthought—using the existing wheels of a wheeled robot is a more elegant way to do it.
Four steerable wheels with in-hub motors provide control in all three axes (yaw, pitch, and roll). You’ll notice that when the robot is tossed, the wheels all toe inwards (or outwards, I guess) by 45 degrees, positioning them orthogonal to the body of the robot. The front left and rear right wheels are spun together, as are the front right and rear left wheels. When one pair of wheels spins in the same direction, the body of the robot twists in the opposite way along an axis between those wheels, in a combination of pitch and roll. By combining different twisting torques from both pairs of wheels, pitch and roll along each axis can be adjusted independently. When the same pair of wheels spin in directions opposite to each other, the robot yaws, although yaw can also be derived by adjusting the ratio between pitch authority and roll authority. And lastly, if you want to sacrifice pitch control for more roll control (or vice versa) the wheel toe-in angle can be changed. Put all this together, and you get an enormous amount of mid-air control over your robot.
Image: Robotics Research Center/West Point
The AGRO robot features four steerable wheels with in-hub motors, which provide control in all three axes (yaw, pitch, and roll).
According to a paper that the West Point group will present at the 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), the overall objective here is for the robot to reach a state of zero pitch or roll by the time the robot impacts with the ground, to distribute the impact as much as possible. AGRO doesn’t yet have a suspension to make falling actually safe, so in the short term, it lands on a foam pad, but the mid-air adjustments it’s currently able to make result in a 20 percent reduction of impact force and a 100 percent reduction in being sideways or upside-down.
The toss that you see in the video isn’t the most aggressive, but lead author Daniel J. Gonzalez tells us that AGRO can do much better, theoretically stabilizing from an initial condition of 22.5 degrees pitch and 22.5 degrees roll in a mere 250 milliseconds, with room for improvement beyond that through optimizing the angles of individual wheels in real time. The limiting factor is really the amount of time that AGRO has between the point at which it’s released and the point at which it hits the ground, since more time in the air gives the robot more time to change its orientation.
Given enough height, the current generation of AGRO can recover from any initial orientation as long as it’s spinning at 66 rpm or less. And the only reason this is a limitation at all is because of the maximum rotation speed of the in-wheel hub motors, which can be boosted by increasing the battery voltage, as Gonzalez and his colleagues, Mark C. Lesak, Andres H. Rodriguez, Joseph A. Cymerman, and Christopher M. Korpela from the Robotics Research Center at West Point, describe in the IROS paper, “Dynamics and Aerial Attitude Control for Rapid Emergency Deployment of the Agile Ground Robot AGRO.”
Image: Robotics Research Center/West Point
AGRO 2 will include a new hybrid wheel-leg and non-pneumatic tire design that will allow it to hop up stairs and curbs.
While these particular experiments focus on a robot that’s being thrown, the concept is potentially effective (and useful) on any wheeled robot that’s likely to find itself in mid-air. You can imagine it improving the performance of robots doing all sorts of stunts, from driving off ramps or ledges to being dropped out of aircraft. And as it turns out, being able to self-stabilize during an airdrop is an important skill that some Humvees could use to keep themselves from getting tangled in their own parachute lines and avoid mishaps.
Before they move on to Humvees, though, the researchers are working on the next version of AGRO named AGRO 2. AGRO 2 will include a new hybrid wheel-leg and non-pneumatic tire design that will allow it to hop up stairs and curbs, which sounds like a lot of fun to us. Continue reading →
#437592 Coordinated Robotics Wins DARPA SubT ...
DARPA held the Virtual Cave Circuit event of the Subterranean Challenge on Tuesday in the form of a several hour-long livestream. We got to watch (along with all of the competing teams) as virtual robots explored virtual caves fully autonomously, dodging rockfalls, spotting artifacts, scoring points, and sometimes running into stuff and falling over.
Expert commentary was provided by DARPA, and we were able to watch multiple teams running at once, skipping from highlight to highlight. It was really very well done (you can watch an archive of the entire stream here), but they made us wait until the very end to learn who won: First place went to Coordinated Robotics, with BARCS taking second, and third place going to newcomer Team Dynamo.
Huge congratulations to Coordinated Robotics! It’s worth pointing out that the top three teams were separated by an incredibly small handful of points, and on a slightly different day, with slightly different artifact positions, any of them could have come out on top. This doesn’t diminish Coordinated Robotics’ victory in the least—it means that the competition was fierce, and that the problem of autonomous cave exploration with robots has been solved (virtually, at least) in several different but effective ways.
We know Coordinated Robotics pretty well at this point, but here’s an introduction video:
You heard that right—Coordinated Robotics is just Kevin Knoedler, all by himself. This would be astonishing, if we weren’t already familiar with Kevin’s abilities: He won NASA’s virtual Space Robotics Challenge by himself in 2017, and Coordinated Robotics placed first in the DARPA SubT Virtual Tunnel Circuit and second in the Virtual Urban Circuit. We asked Kevin how he managed to do so spectacularly well (again), and here’s what he told us:
IEEE Spectrum: Can you describe what it was like to watch your team of robots on the live stream, and to see them score the most points?
Kevin Knoedler: It was exciting and stressful watching the live stream. It was exciting as the top few scores were quite close for the cave circuit. It was stressful because I started out behind and worked my way up, but did not do well on the final world. Luckily, not doing well on the first and last worlds was offset by better scores on many of the runs in between. DARPA did a very nice job with their live stream of the cave circuit results.
How did you decide on the makeup of your team, and on what sensors to use?
To decide on the makeup of the team I experimented with quite a few different vehicles. I had a lot of trouble with the X2 and other small ground vehicles flipping over. Based on that I looked at the larger ground vehicles that also had a sensor capable of identifying drop-offs. The vehicles that met those criteria for me were the Marble HD2, Marble Husky, Ozbot ATR, and the Absolem. Of those ground vehicles I went with the Marble HD2. It had a downward looking depth camera that I could use to detect drop-offs and was much more stable on the varied terrain than the X2. I had used the X3 aerial vehicle before and so that was my first choice for an aerial platform.
What were some things that you learned in Tunnel and Urban that you were able to incorporate into your strategy for Cave?
In the Tunnel circuit I had learned a strategy to use ground vehicles and in the Urban circuit I had learned a strategy to use aerial vehicles. At a high level that was the biggest thing I learned from the previous circuits that I was able to apply to the Cave circuit. At a lower level I was able to apply many of the development and testing strategies from the previous circuits to the Cave circuit.
What aspect of the cave environment was most challenging for your robots?
I would say it wasn't just one aspect of the cave environment that was challenging for the robots. There were quite a few challenging aspects of the cave environment. For the ground vehicles there were frequently paths that looked good as the robot started on the path, but turned into drop-offs or difficult boulder crawls. While it was fun to see the robot plan well enough to slowly execute paths over the boulders, I was wishing that the robot was smart enough to try a different path rather than wasting so much time crawling over the large boulders. For the aerial vehicles the combination of tight paths along with large vertical spaces was the biggest challenge in the environment. The large open vertical areas were particularly challenging for my aerial robots. They could easily lose track of their position without enough nearby features to track and it was challenging to find the correct path in and out of such large vertical areas.
How will you be preparing for the SubT Final?
To prepare for the SubT Final the vehicles will be getting a lot smarter. The ground vehicles will be better at navigation and communicating with one another. The aerial vehicles will be better able to handle large vertical areas both from a positioning and a planning point of view. Finally, all of the vehicles will do a better job coordinating what areas have been explored and what areas have good leads for further exploration.
Image: DARPA
The final score for the DARPA SubT Cave Circuit virtual competition.
We also had a chance to ask SubT program manager Tim Chung a few questions at yesterday’s post-event press conference, about the course itself and what he thinks teams should have learned from the competition:
IEEE Spectrum: Having looked through some real caves, can you give some examples of some of the most significant differences between this simulation and real caves? And with the enormous variety of caves out there, how generalizable are the solutions that teams came up with?
Tim Chung: Many of the caves that I’ve had to crawl through and gotten bumps and scrapes from had a couple of different features that I’ll highlight. The first is the variations in moisture— a lot of these caves were naturally formed with streams and such, so many of the caves we went to had significant mud, flowing water, and such. And so one of the things we're not capturing in the SubT simulator is explicitly anything that would submerge the robots, or otherwise short any of their systems. So from that perspective, that's one difference that's certainly notable.
And then the other difference I think is the granularity of the terrain, whether it's rubble, sand, or just raw dirt, friction coefficients are all across the board, and I think that's one of the things that any terrestrial simulator will both struggle with and potentially benefit from— that is, terramechanics simulation abilities. Given the emphasis on mobility in the SubT simulation, we’re capturing just a sliver of the complexity of terramechanics, but I think that's probably another take away that you'll certainly see— where there’s that distinction between physical and virtual technologies.
To answer your second question about generalizability— that’s the multi-million dollar question! It’s definitely at the crux of why we have eight diverse worlds, both in size verticality, dimensions, constraint passageways, etc. But this is eight out of countless variations, and the goal of course is to be able to investigate what those key dependencies are. What I'll say is that the out of the seventy three different virtual cave tiles, which are the building blocks that make up these virtual worlds, quite a number of them were not only inspired by real world caves, but were specifically designed so that we can essentially use these tiles as unit tests going forward. So, if I want to simulate vertical inclines, here are the tiles that are the vertical vertical unit tests for robots, and that’s how we’re trying to to think through how to tease out that generalizability factor.
What are some observations from this event that you think systems track teams should pay attention to as they prepare for the final event?
One of the key things about the virtual competition is that you submit your software, and that's it. So you have to design everything from state management to failure mode triage, really thinking about what could go wrong and then building out your autonomous capabilities either to react to some of those conditions, or to anticipate them. And to be honest I think that the humans in the loop that we have in the systems competition really are key enablers of their capability, but also could someday (if not already) be a crutch that we might not be able to develop.
Thinking through some of the failure modes in a fully autonomous software deployed setting are going to be incredibly valuable for the systems competitors, so that for example the human supervisor doesn't have to worry about those failure modes as much, or can respond in a more supervisory way rather than trying to joystick the robot around. I think that's going to be one of the greatest impacts, thinking through what it means to send these robots off to autonomously get you the information you need and complete the mission
This isn’t to say that the humans aren't going to be useful and continue to play a role of course, but I think this shifting of the role of the human supervisor from being a state manager to being more of a tactical commander will dramatically highlight the impact of the virtual side on the systems side.
What, if anything, should we take away from one person teams being able to do so consistently well in the virtual circuit?
It’s a really interesting question. I think part of it has to do with systems integration versus software integration. There's something to be said for the richness of the technologies that can be developed, and how many people it requires to be able to develop some of those technologies. With the systems competitors, having one person try to build, manage, deploy, service, and operate all of those robots is still functionally quite challenging, whereas in the virtual competition, it really is a software deployment more than anything else. And so I think the commonality of single person teams may just be a virtue of the virtual competition not having some of those person-intensive requirements.
In terms of their strong performance, I give credit to all of these really talented folks who are taking upon themselves to jump into the competitor pool and see how well they do, and I think that just goes to show you that whether you're one person or ten people people or a hundred people on a team, a good idea translated and executed well really goes a long way.
Looking ahead, teams have a year to prepare for the final event, which is still scheduled to be held sometime in fall 2021. And even though there was no cave event for systems track teams, the fact that the final event will be a combination of tunnel, urban, and cave circuits means that systems track teams have been figuring out how to get their robots to work in caves anyway, and we’ll be bringing you some of their stories over the next few weeks.
[ DARPA SubT ] Continue reading →