Build your physical track for AWS DeepRacer
This section describes how you can build a physical track for a AWS DeepRacer model. To drive your AWS DeepRacer autonomously and to test your reinforcement learning model in a physical environment, you need a physical track. Your track resembles the simulated track used in training and replicates the environment used to train the deployed AWS DeepRacer model.
For the best experience, we recommend using pre-printed tracks and track barriers. Using
pre-printed tracks and barriers facilitates the smooth set up and installation of the
AWS DeepRacer track environment. Instead of building a track from scratch, you assemble pre-printed sections of track and track barriers. When your
event is over, you can disassemble and store and reuse the pre-printed tracks and barriers for future events.
Pre-printed tracks and barriers as well as details for
estimating space and other requirements for events are available at AWS DeepRacer
Storefront.
Topics
