Back in early 2003, I started looking for some work to do over the summer and found out about some interesting research being conducted by one of my teachers. He was working on self-organizing systems where a large number of individuals can organize themselves through local, independent observations, rather than using a system of coordinated, large-scale communications.

In the abstract, his system utilized artificial physics where robots would repel each other when too close and attract each other when far apart, which would eventually stabilize into different patterns depending on the parameters used. To test this in the real world, he used simple kits provided by KIPR that included Legos (to build robot bodies), motors, servos, and small computers to control robot logic.

Although my role in the project was rather modest, the time I spent there was quite enjoyable. I started out running a few baseline tests (which ended up helping us work out a new optical sensor configuration); took a look through the C code controlling the robots; and created a new, more stable base for the robots that used fewer Legos and so allowed us to build more robots from each kit.