It could be a scene from a science-fiction movie: As a cloud of toxic fog slowly spreads through a labyrinth beneath them, ten Rubik’s cube-sized robots roll and turn on a giant plate of glass.

Like a mechanical ballet, they sense the air below and move in unison, working their ways to the edge of the ever-changing cloud. Relaying information on the fog’s position and perimeter, they aid in quarantining and mitigating the spreading toxin.

Robots and fog may create a mysterious or surreal combination, but to Utah State researcher YangQuan Chen, it is the future of homeland security.

Chen, an assustant professor in the Department of Electrical and Computer Engineering, is working on ways to sense and eradicate dangerous substances that could spread on and through air, land, and water. To do this, he has created a simplified simulation of a dangerous event using commercial fog, Plexiglas, two-by-fours, and his custom robots, called MASmotes.

“I play with fog because fog is cheap and harmless,” says Chen. “I do it to find out how other things work.”

Chen’s system is called MAS-net, or Mobile Actuator-Sensor Networks. It is based on the idea that new advances in wireless technology may allow sensors that were previously immobile to move around, based on changing conditions.

In the past, most multiple-input sensors were organized in arrays, or regularly spaced intervals, and they did not have the ability to move to and sense the most interesting or important parts of their environment. Advances in wireless Internet have taken the limits off of these sensors and have opened up new possibilities for sensor research.

“Recently, people have really latched on to the idea that you can communicate and talk to one another using the Internet,” says Chen. “Now, you can communicate with the world, but our question asks whether you can sense the world—and maybe control it a little bit.”

Chen’s vision of the future is the capability not just to prevent some terrorist attacks but to respond to those acts, should they occur. Chen is working to develop technology that can sense the boundaries of dangers such as airborne toxin clouds, nuclear radiation fields, or pollutants in lakes or reservoirs. These dangers spread and travel, and mobile sensors may help to better track and eliminate them.

“Our big picture applies to land, water, and air,” says Chen. “When there is a distributed system that spreads out, such as fog or an oil slick, it must be sensed and measured in both space and time.”

To address all types of toxic terrorist acts, Chen’s MAS-net sensors eventually will be installed on any type of mobile vehicle, including unmanned ground robots, airplanes and helicopters, or autonomous boats.

For now, though, Chen is working with MASmote robots at USU’s Center for Self-Organizing Intelligent Systems. The robot/fog MAS-net test system was actually designed and developed with the help of Chen’s graduate assistants, Zhen Song, Zhomin Wang, Pengyu Chen, and Anisha Arora.

“I provide the big picture and the original idea, but the students work to flesh out different parts of the research,” says Chen.

Chen’s students created an area rug-sized hollow box with obstacles in it to create different types of fog flows and covered it with a transparent top. As air and fog circulate through the box, the ten small MASmotes use their optical sensors to detect the fog beneath. A base station computer, which is connected to the robots with a wireless network and an overhead camera-based pseudo-GPS system that can track their positions, then predicts the spread of the fog, computes new sampling locations, and redirects the robots.

The research team completed the MAS-Net system and entered it in the 2005 Crossbow Smart Dust Challenge, which looks for the best executable ideas for wireless sensor networks. This USU team won second place at the national competition.

Chen, however, has more than MAS-net to thank his students for. His teaching experience at Utah State helped steer him into his current research field. When he received his Ph.D., Chen’s emphasis was iterative learning control, a very specialized engineering field. Though he was at the top of his field, having written the second research monograph on the topic, he realized that this line of research was somewhat limited.

“In Spring 2002, I was teaching a course, ECE 7750, which is on distributed control systems,” says Chen. “This course was never offered before. I was fascinated. It stimulated me, and it interested the students, who enjoyed the challenge of the course. The dynamics of the topic provided enough complexities to really play with.”

Since then, Chen has jumped in head-first into distributive control systems research, and hasn’t looked back.

“No one can force you to do research,” says Chen. “You have to have an inner drive and passion to do it. You have to like it.”

It’s good that Chen likes his research, because there is an exciting future ahead. The MAS-Net platform will provide avenues of high-level research for another 5 to 10 years. That means a lot more foggy forecasts for the robot lab, but it also means a bright future for advances in homeland security measures.

- Anna McEntire