An End-to-End Autonomous Vehicle Trained Entirely in simulation that can Drive in the Real World
Non-Exclusively Licensed
Invention type: Technology
/
Case number: #22034
A controller for an autonomous vehicle is trained using simulated paths on a roadway and simulated observations that are formed by transforming images previously acquired on similar paths on that roadway. Essentially an unlimited number of paths may be simulated, enabling optimization approaches including reinforcement learning to be applied to optimize the controller.
Researchers
Daniela Rus
/
Jacob Phillips
/
Julia Moseyko
/
Alexander Amini
/
Igor Gilitschenski
/
Sertac Karaman
Departments: Dept of Electrical Engineering & Computer Science, Computer Science & Artificial Intelligence Lab, Department of Aeronautics and Astronautics
Technology Areas: Artificial Intelligence (AI) and Machine Learning (ML) / Computer Science: Networking & Signals / Industrial Engineering & Automation: Autonomous Systems
Impact Areas: Sustainable Future, Connected World
-
simulation-based training of an autonomous vehicle
United States of America | Published application
License this technology
Interested in this technology? Connect with our experienced licensing team to initiate the process.
Sign up for technology updates
Sign up now to receive the latest updates on cutting-edge technologies and innovations.