Software and hardware for robotics
Code, create and collaborate with our GitHub community to make your AWS DeepRacer more than the sum of its parts.
Featured projects
Follow the Leader
By AWS DeepRacer team
Learn to deploy an object detection model that enables the AWS DeepRacer device to identify and follow an object. Extend this sample project by modifying the code to recognize other objects for your use case. Bring your own custom model, navigation logic, and add-on hardware (optional) to invent your own application.
RoboCat
By Martin Paradesi, Organizer of AWS Machine Learning Community: NYC
Community project that leverages OpenCV image-processing capabilities to detect a mouse in the image from an infrared camera mounted on the AWS DeepRacer. The infrared camera allows the DeepCat to detect the mouse in the dark and scare it away using a pre-configured action space with 3 actions (move forward if a mouse is detected, move back after previous action, and remain stationary).
Project ideas: Pest control, object tracking, and in-home safety
Mapping
By AWS DeepRacer team
DeepDriver
By Jochem Lugtenburg, Developer at Relive
Mimics a real-world car that starts and stops at traffic lights and stop signs. The logic for identifying different colors in traffic signals and detect stop signs was developed by combining various computer vision capabilities, including OpenCV image- processing functions and object-detection machine learning models. This project was created by a 2019 and 2020 AWS DeepRacer Championship Cup Finalist.
Project ideas: Multi-racer competition (driving slower if another AWS DeepRacer is in front of the car or overtaking other cars using the LIDAR sensor), autonomous vehicle, using AI for safety and perception (training the car to stop when a person or animal walks in front of it and staying within the bounds of a lane)
Off road
By AWS DeepRacer team
Project ideas: Build your own track to race or create fun games (get the AWS DeepRacer to find a way out of a maze)
DeepBlaster
By Chris Miller, CEO of Cloud Brigade
Use object detection models to build a toy that can identify objects and fire balls at them using a motorized toy blaster mounted on the AWS DeepRacer. It extends the AWS DeepRacer hardware with a 3D-printed, servo-controlled rotating turret and custom bracket. An Arduino micro controller is used to control the rotation servo, blaster flywheels, and ammo feeder motor over a USB/Serial link. A second camera can be added to the turret to allow AWS DeepRacer to scan 360 degrees for objects and fire balls at them.
Project ideas: Multi-tasking (navigate and fire a nerf cannon), competitive racing, strategic games
Connect with the community
Get answers to your questions and collaborate on projects.
Look under the hood
AWS DeepRacer
Developers of all skill levels can get hands-on with machine learning with the AWS DeepRacer, a 1/18-scale race car.
AWS DeepRacer Evo
AWS DeepRacer Evo is the next generation in autonomous driving. It comes fully equipped with stereo cameras and a LiDar sensor.