Watch DARPA’s SubT Cave Circuit Virtual Competition

Fully autonomous teams of robots will hunt for artifacts in DARPA’s virtual cave environment

3 min read
DARPA SubT
Image: DARPA

While we’re super bummed that COVID forced the cancellation of the Systems Track event of the DARPA Subterranean Challenge Cave Circuit, the good news is that the Virtual Track (being virtual) is 100 percent coronavirus-free, and the final event is taking place tomorrow, November 17, right on schedule. And honestly, it’s about time the Virtual Track gets the attention that it deserves—we’re as guilty as anyone of focusing more heavily on the Systems Track, being full of real robots that alternate between amazingly talented and amazingly klutzy, but the Virtual Track is just as compelling, in a very different way.

DARPA has scheduled the Cave Circuit Virtual Track live event for Tuesday starting at 2 p.m. ET, and we’ve got all the details.

If you’ve been mostly following the Systems Track up until this point, you should definitely check out the article that the Urban Circuit Virtual Track winning team, Michigan Tech’s Team BARCS, wrote for us last week. It’s a great way of getting up to speed on what makes the virtual SubT competition so important, and so exciting.

All the Virtual Track teams that submitted their code have absolutely no idea how well their virtual robots did, and they’ll be watching their runs at the same time as we are.

The really amazing thing about the Virtual Track is that unlike the Systems Track, where a human in the loop can send commands to any robot in communications range, the virtual teams of robots operate fully autonomously. In fact, Virtual Track teams sent their code in weeks ago, and DARPA has been running the competition itself in secret, but on Tuesday, everyone will find out how they did. Here’s the announcement:

On Tuesday, November 17 at 2PM EST, the Defense Research Projects Agency (DARPA) will webcast its Subterranean (SubT) Challenge Cave Circuit Virtual Competition. Viewers can follow virtual versions of real autonomous robots, driven by software and algorithms created by 16 competitors, as they search a variety of virtual cave environments for target artifacts. The SubT Challenge is helping DARPA develop new tools for time-sensitive combat operations or disaster response scenarios. The winners of this virtual showcase will be announced at the end of the webcast, and $500,000 worth of prizes is at stake.

What we’re really looking forward to on Tuesday is the expert commentary. During past Systems Track events, live streaming video was available of the runs, but both the teams and the DARPA folks were far too busy running the actual competition to devote much time to commentating. Since the virtual competition itself has already been completed, we’ll be getting a sort of highlights show on Tuesday, with commentary from DARPA program manager Tim Chung, virtual competition lead Angela Maio, along with Camryn Irwin, who did a fantastic job hosting the Urban Circuit livestream earlier this year. We’ll be seeing competition run-throughs from a variety of teams, although not every run and not in real-time of course, since the event is only a couple hours long. But there will be a lot more detail than we’ve ever had before on technology and strategy directly from DARPA.

All the Virtual Track teams that submitted their code have absolutely no idea how well their virtual robots did, and they’ll be watching their runs at the same time as we are. I’ll be on Twitter for the entire event (@BotJunkie) to provide some vaguely informed and hopefully amusing commentary, and we’re hoping that some of the competing teams will be on Twitter as well to let us know how happy (or sad) they are with how their robots are performing. If you have questions, let me know, and we’ll do our best to get in touch with the teams directly, or go through DARPA during a post-event press briefing scheduled for Wednesday.

[ DARPA SubT Virtual Cave Circuit Livestream ]

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