Thoughts about the DARPA Grand Challenge…

I’m getting bombarded with emails about the incipient DARPA grand challenge in disaster robots– very exciting- both the idea and the attention rescue robotics is getting!

While I haven’t gone through the BAA in detail (the whole email barrage thing plus I teach on Tuesdays), the media coverage and speculation highlights 3 things that I especially like:

The idea of integration is fantastic and a key enabler in making robots adoptable. Since 1999, we’ve seen this gap between an interesting sensor or mobility platform and the “full meal deal” of working in a scenario.

Another interesting idea is the use of humanoids. Up until Fukushima, rescue robots have been primarily used for sub-human scale space– spaces where people simply couldn’t go because they didn’t fit. Fukushima and indeed chemical disasters such as Bhopal occur in human-scaled spaces, where people can physically fit but may not be able to survive or work long (or well) with protective gear. The rule of thumb is that robots don’t replace people or dogs, they do things that humans can’t do or can’t do for long enough or well enough– hence our name: Center for Robot-Assisted Search and Rescue. Through funding by the National Science Foundation, we’ve been working with TEEX on human-robot interaction for chemical, biological, radiological and nuclear events- and we see huge possibilities for land, sea, and aerial robots.

And it’s win-win: the impact of improved manipulation would benefit robots operating in either scale of space– the skills that allow a a large robot to open doors could be used by a small robot to move rubble out of the way or help triage an unconscious victim.

The focus on the media appears to be on humanoids, which I hope doesn’t detract from other types of mobility or modalities. There are often aerial and water-based aspects of disasters- at Fukushima, Westinghouse used the Honeywell UAV to sample radiation and get close up views of structural damage (I assisted the Westinghouse team). Marine robots could have been used to monitor pollution in the sea.

And keep in mind that from a robotics perspective, there are at least 12 very distinct activities for rescue robots beyond the direct intervention needed to have prevented the explosions as Fukushima. These are search, reconnaissance and mapping, rubble removal, debris estimation, structural inspection, in situ medical assessment and intervention, medically-sensitive extrication and evacuation of casualties, acting as mobile beacon or repeater, adaptive shoring, logistics support, victim recovery, and serving as a surrogate for a team member. This list was compiled based on feedback from responders and what they’ve asked for or speculated on based on our 15 deployments and 30+ exercises we’ve participated in.

A good starting place is Chapter 50 Search and Rescue Robots in Handbook of Robotics and I’m working hard on my forthcoming book on Rescue Robots for MIT Press.