Eric Cawi

For my senior design project at George Mason University, I worked in a team of four students to design an autonomous search and rescue unit. We broke the task up into four subgroups - path planning, mapping, target recognition, and rescue.

 

My module was path planning: Given a probability distribution from my previous SAR research, what is the best flight path for a UAV? We took a recursive approach, starting with a large search area and breaking down into smaller and smaller cells until the unit's camera resolution was achieved. At each point, local hill climbing algorithms are used to generate the best path through each level. Finally, the path is converted to latitude/longitude and input to the flight controller on our quadcopter.