ADAS Tools for Verification, Validation and Development in Synthetic Environments

Project Team

Principal Investigator

Bogdan Epureanu, University of Michigan

Government

David Gorsich, Mark Brudnak, John Brabbs, U.S. Army GVSC

Charne Folks, U.S. Army Aberdeen Test Center

Industry

Kim McGarrity, Dave Hoadley, Jason Moore, Arnie Berlin, Bill Maher, MathWorks

Reid Steiger, Peggy Caveney, Emily Elazzazi, Chris Davey, Ford Motor Company

Student

Samudhbhav Prabhu Srinivas, Ted Sender, Aabhaas Vaish, Igor Khalip, Jack Decker, University of Michigan

Project Summary

Project began in 2020 and was completed in 2021.

The urgent modernization needs of the U.S. Army require new approaches to enable fielding technologies at a faster pace. Fast technological adoption requires new approaches to design, test, validate and develop complex systems. In particular, prototyping and experimentation must be done at a faster pace than ever before. However, prototyping and experimentation are slow and costly especially when used iteratively in technology adoption and design processes.

One approach to reduce the number of physical iterations is to perform them in synthetic environments. This approach will minimize the need to build physical prototypes. Such virtual prototyping can reduce risks by allowing failure at low cost. The ultimate aim is therefore to enable virtual prototyping combined with methods to evaluate operational effectiveness through modeling and simulation to enable faster fielding of technology with a clear understanding of the operational capability of the technology. However, creating synthetic environments requires novel modeling and simulation capabilities.

The main goal of this project is to create autonomous driving tools by using synthetic environments that will enable virtual prototyping of algorithms, methods and technologies for autonomous vehicles. This includes the development of software tools for sensor validation, scene generation, and methodologies to review logged sensor data in a virtual environment after the data has been collected and/or processed.

#5.A71