Trevor Smith

BIO

Trevor Smith graduated from WVU in 2021 with a BS in Mechanical Engineering and a BS in Aerospace Engineering. Now he is pursuing his PhD in Mechanical Engineering, studying computational swarm robotics. Outside of research he enjoys robotics competitions and outreach. He competes in the University Rover Challenge Team as an Overall Team lead and as and inter-subteam. In addition to being a member of the WVU Robotics club (that teaches primarily freshman and sophmores about robotics), he mentors local high school Vex Robotics teams.

RESEARCH STATEMENT

Robotic swarms are a newly emerging field in robotics that provides the advantages of high adaptability and non-single point failure, over conventional single agent robots. Since robotic swarms leverage simple and local agent rules, they can achieve tasks that are intractable for conventional robots to solve. This is because each robot in the swarm only has to solve a simple problem, which collectively emerges to solve the complex global task. While conventional robotics require an individual robot to directly solve the entire complex problem. This process perfectly mirrors computational methods such as Computational Fluid Dynamics (CFD). Where instead of trying to analytically solve a complex flow field (conventional robotics), the flow is discretized into multiple simple fluid elements that each solve for the local flow at that element (robotic swarms). Therefore, if each robot in the swarm is envisioned as a finite element, computational methods, such as CFD, can be applied as agent level rules. This mitigates the great difficulty of determining appropriate local agent rules to achieve a given global task, due to the implicitness of global swarm robotics solutions. Thus my research is in developing a generalized method for designing swarm robotics algorithms using computational methods