Meeting with Professor Azad 09/09/22

Picture of the autonomous go cart.

09/09/22

FYDP meeting with Professor Azad

Attendees: Riyad Khan, Jorge Castro, Shivam Bhatt, Joshua Unrau, Professor Nasser Azad, Masters Student

Pre meeting notes

Notes:

• Our goals are to work on something that is autonomous, useful, challenging, exciting and preferably novel.

• A good strategy to pursue is to find some novel or niche problem space or area and tackle it in a unique way to make us stand out from other projects or already existing implementations.

• Current philosophy is to try to work on something both achievable and ambitious. The key here will be to have the ability to properly scale the solution and scale back if necessary or pressed for time.

• Joshua is interested in problem spaces involving biological systems such as modeling the Human body skeleton and muscle model for predicting injuries and fractures combined with sensors.

• Other search spaces could include problems related to using automation in underground mines.

Actual Meeting notes

• Professor Nasser was very friendly and talked about many different ideas. He was very enthusiastic about autonomous systems and reinforcement learning applied to control.

• Autonomous navigation of crowded airports

• Autonomous lane changing

• Multi-agent systems and follower robots

• Use of reinforcement learning for control applications

• We got to see the autonomous golf cart created by ClearPath robotics larger than we expected. It was much bigger than the one Jorge worked on. The cart has lidars, gps, imu, cameras and a gps corrector station.

• We also were able to see the smart city roadway that was constructed for testing autonomous systems. This lab is being moved to one floor below.

Post meeting notes

• We all agreed that the problems related to the actual cart itself are all too software heavy as it already has many electrical and hardware components already.

• There is potential for vastly improving the smart city lab, especially with regards to adding verticality to challenge testing of flying drones like the quadcopter. Generally building a scaled environmental modal may be a good idea.

TODOS

• Look around for sponsorships or ideas from other companies or past coops

• Not entirely necessary but we could also consult with other professors specifically MTE 544 Yue Hu and ECE 484 Michael Fisher

Some images of the smart city lab