We’ve heard from some of our colleagues in the field about rescue robots but now the issue has shifted to nuclear response… I’m getting lots of inquiries.

About the rescue robots: The rescue roboticists had contacted various fire departments who for this phase did not need the ground robots (see previous blogs) as the tasks are getting people off the tops of buildings or trying to recover bodies. However, we have reports that there is considerable interest in robots for recovery- especially inspecting port and underwater infrastructure as well as in removing rubble. I have no confirmation that they have deployed any robots for these tasks. We have shifted our standby cache to include more underwater vehicles with very accurate sonars.

Now to nuclear response–

Red Whittaker at the Field Robotics Institute at CMU is the expert in using robots for nuclear disasters (had robots at the Chernobyl and Three Mile Island- but many months later). A few years ago CRASAR looked at what it would take to use our small rescue robots to search for survivors in the aftermath of a “dirty bomb” where radiation wouldn’t be as intense (the scenario from The Sum of All Fears)- and even for this “easy” case, it  wasn’t feasible.

Sensors would probably be the first to go– video and cameras are fairly sensitive to radiation from their CCD chips. It’s impossible to work remotely if you can’t see.

The bigger, slower bomb squad robots were first invented by Oak Ridge National Laboratories to handle nuclear disasters which spun off Remotec. These robots have to big to carry all the shielding. The newer ones are lighter and less protected and the IED robots have evolved to be even lighter- so less reliable in a nuclear disaster. So in some sense you need a dinosaur robot- big, beefy, slow, and stupid (as in few processors)– and even then it’s just a matter of time before enough radiation fries something important… You don’t know how long you’ve got.

Big and stupid means slow. And limited battery times- and who will be changing those batteries? You have to go in and out… losing time at each “lap.”

Stupid is the wrong way to go based on our human-robot interaction studies. Less sensors, particularly cameras means harder to control or move quickly. And less sensors means no autonomy so if you get tired, the robot runs into things. And that could make things worse. Or ruin the robot.  So you want more autonomy so that the robot drives itself, much like a horse. The person directs but the horse actually controls its own motions and adjusts it gait and goes around obstacles.

And then there’s the issue of using a tether or wireless– if radiation doesn’t interfere with wireless, what’s left of the walls and the various containment structures will.

Our hearts and prayers go out to the Japanese. We are, of course, worried about our colleagues in Sendai which is 55 miles from the reactors. There’s no fuel to evacuate.

I’ve been asked by Erico Guizzo, robotics editor at IEEE Spectrum, asks “have robots been used in previous quakes?” The answer is “yes”- just one,  the 2010 Haiti Earthquake.

The US Army Corps of Engineers used a Seabotix ROV (remotely operated underwater  vehicle) to investigate bridge and seawall damage as part of the US assistance to the Haitian government.  An aerospace company from the US self-deployed and flew a small fixed-wing UAV to get an overview of the damage near an orphanage  in ignorance of the Haitian airspace prohibition on all UAVs (the Predator views you saw were taken from outside Haitian airspace)- which caused a bit of a flap (no pun intended).  Note: CRASAR offered (as always, at no cost through our Roboticists Without Borders program) small AirRobot and ISENSYS helicopters for rapid assessment, a VideoRay ROV, and a AEOS water Surface Vehicle with a special sonar particularly well suited for bridge inspection in shallow water, and our usual ground vehicles suited for commercial building collapses (not a lot in Haiti)   to support the response through US Southern Command but was declined by the Haitian officials who said with ample justification that there were too many responders and NGOs pouring in (at that point many were self-deployed, which has the unintended consequence of saturating the officials and causing them to say just say no).

Prof. Daniele Nardi at Sapienza University of Rome, who is one of the leading European researchers in rescue robotics, demonstrated  his micro-quad rotor UAV in the aftermath of the 2009 L’aquila earthquake (I got to attend!), but that was intended for the Fire Service to evaluate the device and not a part of the actual response.

Dan Goldman’s work in duplicating sandfish made it into Science News– a reminder of how biomimetic robots could make a real difference in disasters such as the New Zealand earthquake with its dense rubble. The responders continue to find survivors which is fantastic but are racing the clock. We wish we were there to help with more than our prayers- may all the survivors be found quickly, families reunited, and the recovery be quick!

Juan Rojas saw this about the severe flooding in Australia. Small unmanned marine vehicles, both surface (boats) and ROVs (underwater), can be of assistance in inspecting bridges for underwater damage or debris posed to crash into the substructure and damage the bridge– even in high currents and turbidity which prevent divers from manually inspection. CRASAR has used a AEOS USV at Hurricanes Wilma and Ike to inspect bridges and also experimented with ROVs and UUVs. Through funding from the National Science Foundation, we have a AEOS (called the Sea-RAI for Sea Robot-Assisted Inspection) on the ready for research and response.

Our UMV team (Robin  Murphy, Eric Steimle, Michael Lindemuth, David Trejo, Michael Hall, Daryl Slocum, Stefan Hurlebas, Zenon Medina-Cetina)  has an invited paper coming out shortly in the Journal of Intelligent and Robotic Systems on robot-assisted bridge inspection. A short version is here.

CRASAR’s connection in China, Dr. Juan Rojas, saw this about the Bello, Colombia landslide. Our sympathy and prayers for the victims and their families.

Landslides are tough for robots. CRASAR’s Sam Stover and I assisted with the 2005 La Conchita, California, mudslide which is described here. The short version of what we learned was that the land or mud fluidizes and becomes like water, filling every possible gap. Voids are rare and it’s not a matter of a robot penetrating dense rubble, it’s a matter of being able to be an earthworm and burrow into solid dirt.  “Regular” robots can help search collateral damage but aren’t going to be much use for the buried portions.

BUT GREAT POSSIBILITIES! After my Gegenheimer Lecture at Georgia Tech last week, I met with Dr. Dan Goldman who is working on robotic versions of sand lizards which can burrow. This was the first realistic hope I’ve seen for a robot that could quickly penetrate the soil and perhaps find one of those few voids with trapped air (and survivors). Plus Dr. Carolina Chang got involved in rescue robotics after the landslides in her country, so with this kind of focus perhaps we can improve the odds in the future.