Published on July 26, 2021
Written by theregister.com
Legendarily loopy US military (and now also non-military) ideas factory DARPA has launched a $1m competition for underground robots. This September, the SubT Challenge will pit eight teams against each other in a series of tests in the Louisville Mega Cavern, deep under the surface of the US state of Kentucky.
Sadly, despite the number of entrants lending itself to a one-on-one elimination contest, culminating in a mighty fight to the death between the last two ‘bots standing after destroying their foes in a traditional knockout structure, all eight teams will compete alongside each other in tests to detect certain items and situations vital in rescue work.
Each team will get the opportunity to look for backpacks, phones, gas leaks, and trapped survivors, all important signifiers in search-and-rescue operations. After all of the teams have had the chance to detect their targets, the scores will be totted up and the cash doled out to the winning team.
The lack of a combat element seems like a bit of a missed opportunity, especially given the impressive end-of-level boss cavern environment, but DARPA presumably knows what it’s doing.
The SubT Challenge is the most recent in a series of ‘bot-baffling competitions that have seen the world’s favourite unhinged defence agency attempting to get the world’s roboticists to raise their game and move the industry onwards.
The contest, which has been running since 2018, “seeks novel approaches to rapidly map, navigate, and search underground environments during time-sensitive combat operations or disaster response scenarios,” according to the challenge’s official website.
Such search-and-rescue skills are clearly very much required, as recent events in Miami and elsewhere have demonstrated.
The variety of environments which the teams face – including tunnel, urban, and cave landscapes – mean that teams do not have to stick to just one form of robot platform. Indeed, the organisers promote a diversity of thinking, shapes, and platforms to cope with the different scenarios.
“There’s no rules,” said Carnegie Mellon University’s Matt Travers, a leader of the Explorer team, in an interview with The Engineer. “You can do whatever you want. You can restart every year. DARPA doesn’t specify rules in terms of the platform, the strategy, whatever, it’s just ‘here is the playing field, go out and figure out how to do it’.”
“We have three conventional ground vehicles,” Travers continued. “Then we have one legged robot, six drones – four of which are homogeneous – and that’s pretty much the overall system. So nominally about 10 platforms, aerial and ground, and some combination thereof goes out and tries to compete in the competition.”
As the video from last year’s competition [below] shows, that diversity leads to some pretty weird and wonderful-looking robots. At least some of which could probably be adapted for fighting. If the opportunity arose.
“The Grand Challenge model… is really about planting the seed and oftentimes rallying a community together,” said Dr Timothy Chung, programme manager at DARPA’s Tactical Technology Office (TTO), to The Engineer. “Many of the teams… are releasing a lot of their code, as well as their datasets, all open source, so that the community, everyone rises up with access to the latest and greatest. So even if they’re competitors, you still find a lot of co-operation happening.”
Co-operation, huh? We will look forward to the tag-team combat element next year.
See more here: theregister.com
Header image: The Robot Report
Thanks to: https://principia-scientific.com