May 10, 2021
Peraton Employees Race Virtual Cars to Master Leading-Edge Technologies
Peraton won second place overall in the 2021 AWS Baker Tilly DeepRacer League competition, a local league within an international competition where coders train machine learning models to race a car around a virtual track.
There were three races in January, February, and March, each consisting of time trials with multiple different race cars. There was also a physical race with cameras and sensors on the vehicle.
The Peraton team experimented with different strategies to optimize the race car most efficiently: setting waypoints, staying within the track lines, turning smoothly, avoiding collisions, and prioritizing speed. The team used a reinforcement learning model: as the team ran simulations, they defined certain parameters in a reward function to train the model to continually improve its performance.
Virtual racecar driving is a fun and dynamic way to learn about machine learning (ML) and artificial intelligence (AI) capabilities and gain new AWS and cloud computing skills. Peraton is dedicated to train employees in valuable technical skills and nurture career development. The DeepRacer League competition—which also had lessons and symposiums—is just one example of this effort.
During a recent reorganization in early 2021, Peraton created a technology and engineering (T&E) office under Chief Growth Officer (CGO) Mike King to consolidate corporate-level technology functions. The CGO T&E is considering participating in similar training events and making other investments in ML and AI. These technologies are on the leading edge and a focus area for development in the government.
Peraton is currently working on several independent research and development (IRAD) projects that have ML and AI components—gaining new experiences with these technologies are mutually beneficial for Peraton customers and employees.
The Peraton DeepRacer team was comprised of employees across sectors, business units, and office locations, which demonstrates the broad implications of ML and AI technology. ML and AI help shorten decision-making cycles and glean intelligence out of large volumes of data. This is helpful for any analysis, from orbital calculations to cybersecurity threats to border activity.
“It’s critically important for Peraton employees to stay abreast of leading-edge technologies to supply innovative solutions to our customers now and into the future,” said Greg C., Director of Operations at Peraton’s Blacksburg, VA office, who led Peraton’s DeepRacer participation. “We are really proud of the team for their enthusiasm and engagement—and of course, for their win.”