On May 22, in Montreal, within the ICRA2019 conference, the AI Driving Olympics final was held. The goal of this competition is to compare solutions of using AI technologies in autonomous driving. The team JBRRussia representing JetBrains Research became the absolute champion.
The competition was based on the Duckietown platform, which is a reduced model of the urban transport environment, including roads, vehicles, traffic lights, buildings, etc. Each solution has been tested in a simulator and on a real mobile robot in physical environment; Competition included the following disciplines:
- Autonomous driving in the lane, according the traffic rules;
- Driving in a lane with other vehicles
- Driving around the city, with other vehicles including maneuvers on crossroads
Since February, a joint team of students (from Computer Science Center, LETI, HSE, ITMO), with the support of JetBrains Research, has been working on its solution, trying different approaches and improvements. As a result, the solution showed a good result both in the simulator and in the physical robot has been introduced.
- Konstantin Chaika - LETI
- Nikita Sazanovich - HSE
- Egor Zamotaev - ITMO
- Rami Naim - ITMO
- Oleg Suzdalev - ITMO
- Polina Noskova - ITMO
- Artem Ioselevskiy - CSC
- Maxim Scherbakov - CSC
- Alexander Karavaev - ITMO