Title

Integrating Optimized High-Speed Autonomous Control Systems

School Name

Governor's School for Science and Mathematics

Grade Level

12th Grade

Presentation Topic

Computer Science

Presentation Type

Mentored

Abstract

While autonomous vehicles are growing in popularity, there exists a tradeoff between control and speed. With control directly affecting safety of a vehicle, it is prioritized at the detriment of speed. However, with speed being essential to emergency responses, methods are required for optimizing speed while retaining a high degree of control. This research project aims to convert a proportional integral derivative controller, visual servoing system, and potential field navigator into Robot Operating System (ROS) nodes, integrate them with the Rapid Autonomous Complex-Environment Competing Ackermann-steering Robot (RACECAR) developed at the Massachusetts Institute of Technology, tune them for speed, and integrate them together to complete a test course of various obstacles. Our optimized algorithm variants achieved an increase in processing speed by 20 times, but did not outcompete completely rewritten variants in the test courses. This indicates that the optimization methods provide benefits, but that algorithms themselves must be rewritten to operate at the highest efficiency. Future research will incorporate the optimization methods into the rewritten algorithms to benefit from both effects. Additional algorithms will also be considered for revision and optimization.

Location

Neville 206

Start Date

4-14-2018 9:45 AM

Presentation Format

Oral and Written

COinS
 
Apr 14th, 9:45 AM

Integrating Optimized High-Speed Autonomous Control Systems

Neville 206

While autonomous vehicles are growing in popularity, there exists a tradeoff between control and speed. With control directly affecting safety of a vehicle, it is prioritized at the detriment of speed. However, with speed being essential to emergency responses, methods are required for optimizing speed while retaining a high degree of control. This research project aims to convert a proportional integral derivative controller, visual servoing system, and potential field navigator into Robot Operating System (ROS) nodes, integrate them with the Rapid Autonomous Complex-Environment Competing Ackermann-steering Robot (RACECAR) developed at the Massachusetts Institute of Technology, tune them for speed, and integrate them together to complete a test course of various obstacles. Our optimized algorithm variants achieved an increase in processing speed by 20 times, but did not outcompete completely rewritten variants in the test courses. This indicates that the optimization methods provide benefits, but that algorithms themselves must be rewritten to operate at the highest efficiency. Future research will incorporate the optimization methods into the rewritten algorithms to benefit from both effects. Additional algorithms will also be considered for revision and optimization.