Today, January 17th, is the anniversary of the 1995 Kobe, Japan, earthquake which killed over 6,000 people. I am in Japan, ironically, accepting the Motohiro Kisoi Award for academic contributions to rescue engineering. I am thinking about Haiti, what is and what might have been.
My Japanese hosts at the International Rescue Systems institute are sanguine about the long delays in getting new technologies from the labs and into the hands of the responders, agencies, and the victims themselves. It takes time, they say. They are patient; it took 5 years to rebuild Kobe with the resources of one of the great nations of the world.
During my visit this week and through out my meetings with industries and agencies over the years since 9-11, I have been repeatedly struck by the brilliance and potential of many innovations. Unmanned vehicles and sensors offer viewpoints that humans and dogs can’t provide. A dog can tell if a survivor is inside a collapse structure, but not whether it is safe for a rescuer to enter or how to best shore it up or remove rubble without causing further injuries. Wireless communications improve daily.
The field of social computing is a truly new vista- and if harnessed could not only help the victims help supply the responders with information but be more able to help themselves. I marvel at the ideas of Gloria Mark at UC Irvine and Leysia Palen at Colorado for crisis informatics. One of the lessons of the Kobe earthquake was that centralized, official response teams were insufficient and in the end, the volunteers from the survivors were what made the difference in rescuing people and putting out the fires.
I have also been struck by the price we pay as a society by not incorporating more computer science into engineering and business. Some of the most magnificently agile robots have been rejected or failed in the field because the designers had provided a poor human-machine interface- topics taught in computer science but generally neglected in other disciplines. Other innovations have fallen short because the designers did not understand the domain- how the responders actually work- and did not understand how to understand a domain- again something that computer science addresses daily in human-centric design methodologies.
And as can be seen from the news reports from Haiti, the potential contributions of computer science to intelligent, distributed decision making are important too. There is always a tendency to try to make things more centralized, imposed more control and enforced coordination. And the ShadowBowl exercises run by Eric Rasmussen, now at InSTEDD, show that this just doesn’t work. There are too many agencies, NGOs, and emergent groups that may not even know about each other. A few years ago in an outbrief of a major international disaster, I overheard one agency complain bitterly about Doctors Without Borders, which is normally not equated with Satan. As an outsider, it was clear that both agencies had been doing their jobs but had no way of even knowing that the other agency was in the area or opportunistically coordinating activities. The cognitive sciences can help if they’re allowed.
Great progress is being made by viewing these events from what David Woods at Ohio State calls a polycentric viewpoint- that there will always be multiple centers of decision making and the decisions will change in scope and function over time. Rather than try to continue to dynamically create and maintain a top-down structure, embrace the distributed nature of incidents and create architectures that let the groups go about their business while minimizing conflicts, duplication of resources, etc.
I also believe it is also important for scientists to understand context and policy. The US teams sent to Haiti by FEMA can’t use ground robots because they aren’t approved yet. Military UAVs, so popular in Iraq and Afghanistan, must conform to FAA and FCC requirements to be used by civilians. I believe we must begin to educate our students on policy- what that means for funding and adoption and how to positively influence it. A bright star in this regard has been Henrik Christensen’s leadership through the CCC in creating the US Congressional Caucus on Robotics.
I continue to pray for the people of Haiti and the international response community, but I’m slipping in a few prayers for us to change the way we think about technology, computer science, and science education. I am not sanguine, I believe that we’re running out of time.-Prof. Robin R. Murphy, PhD Raytheon Professor of Computer Science & Engineering
