FACULTY

• Devin Balkcom Ph.D
• Laura E. Ray Ph.D

  STUDENTS

• Anne Loomis
• Joshua Pyke
• Nelson Rosa, Jr.

The goal of the robotics project is to provide machines that disaster-response teams find useful. Exploring collapsed buildings, sites of industrial accidents, natural-disaster areas, and battle zones is dangerous and difficult. Machines should increase rescuers' situational awareness, allowing them to work more effectively with less risk. Balkcom and Ray are exploring solutions to two key problems: sensor delivery, and automated physical reasoning about disaster situations.

Sensor technology, including cameras, microphones, gas sensors, and heat sensors, can expand a responder's awareness, helping to locate victims or warning of dangerous situations. The primary challenge is delivering the sensors to the most relevant area. Mobile robots provide one approach; robots may be able to squeeze into tight areas, and are somewhat expendable.

However, even the best-designed robots get stuck when the unexpected happens. The primary focus of the current work on sensor delivery is in freeing stuck vehicles using clever steering or dynamic rocking. Such non-traditional locomotion strategies are used by human drivers when a car is stuck in snow or sand, and a better understanding of these strategies may allow mobile robots to become more capable without redesign.

Even if no survivors are found, search teams are exposed to great physical risks. One of the most common risks is unstable piles of rubble. Balkcom and Ray are exploring the problem of automated reasoning about stability. If the current exploratory work is successful, future projects might include such technologies as "magic" glasses that can warn the wearer of dangerously unstable blocks that should not be moved or walked on.