The second Teacher Week of Race4Scale was held by Russian project partners on April 27th. The day was divided into two parts. Kudrovo presented on the topic of how to prepare for future careers: Traditions blended with innovations, young technicians, robotics and artificial intelligence in school and the making of the Kudrover. ETU “LETI” presented on the topic of simple algorithms of complex learning: Duckietown platform, self-driving cars piloting, digital model of optical sensors and sharp-eyes driverless vehicle.
Driverless vehicle programming and piloting
Intro and exercise presented by Anton Filatov
The Duckietown platform is an open-source project designed to simulate the urban environment and the behavior of unmanned vehicles. Such a platform has been deployed in the laboratory of Saint-Petersburg Electrotechnical University, and we are ready to show everything what it can do. We will tell you what the main tasks are solved by unmanned vehicles, what they are made of, and for which tasks the human intervention is still required. And, of course, in the laboratory you can see with your own eyes how this works on real mobile robots.
Sharp-eyed driverless vehicles
Research project introduced by Maxim Dobrokhvalov
It is clear that self-driving cars are highly demanded in the nearest future. The first task that every unmanned vehicle should solve is to recognize every object in it’s area of view. Students of SPBETU work on the algorithms that would help these cars to see the environment around them.
Digital model of optical sensors
Research project introduced by Dmitry Ivanov
When a researcher develops a new algorithm, he wants to test it on the real data. Usually, he should find any existing dataset that has been captured by completely different people for different purposes. Students of SPBETU managed to implement a tool that allows to model real sensors and to generate data from these sensors. These data might be used by any researcher that works with a particular sensor.