October 29, 2019

AI Race Cars @ Forge DevCon 2019 with AWS DeepRacer!

deepracer

In partnership with AWS, this year at Forge DevCon you will have the chance to drive an AI-powered car on an the official DeepRacer track, so Developers! Start Your AI Engines! 

 

What is AWS DeepRacer?

AWS DeepRacer is a global racing league where developers program and test digital twin versions of the DeepRacer race car. These digital twin algorithms are then used on a physical track with an actual car and used to compete against developers across the globe.

This is a great way to get hands-on with reinforcement learning or RL and experiment with machine learning, all through autonomous driving. You can get started with the virtual car and tracks in the cloud-based 3D racing simulator, and for a real-world experience, you can deploy your trained models onto AWS DeepRacer.

 

So what's happening at Forge DevCon?

Forge DevCon and AWS are partnering to host a full league version of the DeepRacer track on Monday November 18. Stop by to race, cheer for a team, play with a digital twin of the race car in a Forge Viewer, or just talk to AWS and Forge to learn more about machine learning and digital twins.

 

How can I race on the DeepRacer track at Forge DevCon?

At Forge DevCon, we'll have a track and you will be able to compete against other racers if you come prepared with your algorithm. Attendees are invited and encouraged to race on digital twins at the event. All you need to do to participate is to create your own artificially intelligent (AI) race car model using reinforcement learning! You can train and test the car on a simulated track, which is a digital twin of a real race track that will be at the AWS booth at Forge DevCon 2019.

Come stop by and race your car to compete for prizes and the admiration of your peers.  

No experience with AI/ML or coding is required! Just follow the easy self-guided tutorial linked below to create, train, and test your car's racing algorithm on a virtual racetrack.  When you arrive at Forge DevCon, Amazon will be set up in the exhibit area with DeepRacer cars and a race track. You will be able to stop by the booth, upload your model to a car and test your algorithm in real life on a real track. See the section below for more details.   

AWS DeepRacer: Driven by Reinforcement Learning

You will need to do a training to be able to compete and create your own AI models for the use on the DeepRacer cars. 

track

This course was designed by AWS to teach you how to build your own AI models using the Reinforcement Learning principal of AI, read more about it here. The course starts off by orienting you to AWS DeepRacer before diving into RL. You’ll have several opportunities to build, train, and evaluate your own RL models, and then deploy them to the physical car. By the end of the course, you’ll have your own optimized RL model that you can enter into a virtual or physical circuit of the AWS DeepRacer League.

 

Image result for aws deepracer diagram

 

Make sure you bring your trained AI model in an USB stick to be able to load it into the cars we will have at Forge DevCon this year! We will have prices for the winners and a nice dynamic to compete in pairs like Race championships like Formula 1 and Nascar do it! So prepare your AI models and let's get those tires burning! 

Posted by

Jaime Rosales

Senior Developer Consultant

Jaime Rosales is a Senior Developer Advocate with Autodesk Forge. He’s part of the team that helps partners/customers create new products and transition to the cloud with the use of Autodesk’s new Platform - Forge. He joined Autodesk in 2011 through the acquisition of Horizontal Systems the company that developed the cloud-based collaboration systems—now known as BIM 360 Glue the...

Related Posts