In the aftermath of a disaster—an earthquake, fire, building collapse, or terrorist attack—it is often dangerous or even impossible to collect needed information by hand. A Cornell team is creating an automated self-configuring remote-sensor network to do the work.

The idea grew out of studies of how such networks could be used on the battlefield, says Electrical and Computer Engineering Professor Stephen Wicker, who heads the team that includes molecular biologists, device physicists, telecommunications engineers, information and game theorists, and civil engineers.

Initial research is aimed at the detection of biohazards, but the underlying principles can be applied to other situations such as searches for earthquake victims using audio and body-heat sensors, monitoring of municipal water systems for leaks or contamination, and detecting toxic chemicals, extreme heat, or radiation.

The self-configuring, rapidly redeployable wireless bio-sensor network consists of a large number of tiny low-power sensors that can be randomly spread across an area to sample the environment, communicate with one another, and report information to a central location using tiny radio transceivers. Development of the actual sensor devices, the selfconfiguring network, and the operating system are all parts of the project.

Network infrastructure can be tested by building a software simulator to model the network’s behavior in various circumstances. But this approach is slow, and it precludes the design of a network where simulator feedback can be used to dynamically adapt the network to performance bottlenecks or even denialof- service attacks—which is especially important in ad hoc wireless networks where key network properties such as connectivity and channel characteristics vary frequently.

Professor Rajit Manohar in Electrical and Computer Engineering heads a group that is designing an asynchronous VLSI (Very Large Scale Integration) simulator architecture that overcomes these limitations. His “network on a chip” uses an energy-efficient processor that uses less than a microwatt of power and clockless circuits: It is suitable for a large-scale sensor network.

Manohar is co-founder of Cornell’s Computer Systems Lab, where faculty members in the School of Electrical and Computer Engineering conduct research with faculty in the Department of Computer Science on topics in computer architecture, parallel computer architecture, operating systems and compilers, digital and analog VLSI design, and system specification and verification.

The use of small, inexpensive Earth-orbiting satellites in a distributed group can yield high-precision performance that is not possible for a single larger one, while also providing benefits such as increasing reliability and the ability to upgrade technology. New applications for these satellites include advances in space-based interferometry, radar, and global telecommunications. The use of small autonomous airplanes is envisioned for a wide range of applications because humans can be removed from dull or dangerous activities.

Professor Mark Campbell in Mechanical and Aerospace Engineering works on control and autonomy for high-performance systems of spacecraft and aircraft.

He designed and built one of the smallest self-propelled satellites for use in a distributed satellite test bed in space—1.5 ft wide and 1.0 ft tall. Campbell is developing planning, control, sensing, and multiple-agentbased software architectures for clusters of autonomous satellites. In conjunction with industry leaders, he is developing and implementing onboard autonomy, cooperation, and operator decision modeling for multiple aircraft in defense and commercial applications.