Agent-based simulations are a great way to simulate large complex problems that involve multiple interacting things. Human crowds are a perfect problem for agent-based simulation analysis, with multiple independent agents (humans) all with their own unique goals and knowledge interacting within a small space. Indoor disasters such as fire have resulted in countless tragedies throughout history, but the interesting thing about such problems is that often injury and death are more a result of the crowd reaction, rather then the disaster itself. My colleagues and I developed a simulator to model different enclosed spaces and their geometries, disaster scenarios, and the human agents reaction to these scenarios. The simulator verifies that many counter-intuitive room-designs actually do help save lives in the event of a disaster.
Needles are used in the medical world to perform a wide variety of procedures. However, a problem arises when trying to reach a target deep within someones body, the needle begins to bend and the target is missed. Usually this bending is a bad thing, but if we really understand it and can control the bending perhaps it can be used to our advantage. I took part in some cutting edge research being done in the Johns Hopkins University Haptics Research Lab in modeling, controlling, enhancing, and creating a user control interface for this bending effect.