AWS DeepRacer- Beginner Machine Learning Workshop

Makerspace Adelaide in partnership with Amazon, invite you to join us for an exciting two days of artificial intelligence and machine learning leveraging the AWS DeepRacer platform.

In this workshop, we will give you the training and guidance you need to successfully code, train and deploy your own Machine Learning (ML) model for a real-world race car that navigates the racetrack under time trial conditions.

By the end of this workshop, you will have:

• Created a fully autonomous race car model.
• A working understanding of how ML works, and how you could leverage it for your own projects.

Set over 2 days, participants will gain a working knowledge of the AWS Console, AWS DeepRacer, Artificial Intelligence (AI) & Machine Learning, and applying AI & ML to solve real-world problems for their personal projects.

There is no cost to participate in this activity. You will be provided with free, temporary AWS Accounts to use on the first day.

You only need to bring:

1. A reliable laptop & Charger.
2. Yourself

Sessions Breakdown

Participants must attend both sessions to be able to complete this workshop.

Session 1 – Saturday June 5th 2021, 10:30am to 1:00pm

• 10:00 Greetings, briefing, general theory.
• 11:00 Building your first AWS DeepRacer model.
• 13:00 Lunch arrives, final 30 mins of AWS Supported model training.
• 13:30 Finish Session 1.

We will first go through the basics of Machine Learning on AWS and how to use the AWS Console for AWS DeepRacer projects. This will cover the basic theory of AWS DeepRacer, how the car works, what data is exposed for developers to utilize in training their vehicles to self-drive, and some simple reward function theories that could be used as a base for your own model development.

Overnight you can continue to train your model/s in the AWS Console to create and identify the best autonomous racer possible, but remember it is possible to over train your model!

Session 2 – Sunday June 6th 2021, 1:00pm to 4:30pm

• 13:00 General Discussion, Open Q&A as final training opportunities complete.
• 15:45 Virtual Race closes, winners announced.
• 16:00 Next steps, using your AWS Credit, Submitting to Global Competitions
• 16:30 Session 2 Finish

We will begin the day with an open discussion on how everyone went overnight, and there’s just enough time to get 1 or 2 more iterations of your model into the race! With half an hour to go we will conduct the virtual race with the winner being crowned on the day.

Participants who attend both days will receive $75 AWS credits to allow them to continue training DeepRacer models and enter the global competition (if they wish), and you will be guided on how to export your models and reward functions from your temporary AWS accounts provided.

Who should attend?

Anyone with a curiosity towards what is Artificial Intelligence and how to create a Machine Learning project with little to no experience needed.

Prerequisites

You will need a working laptop, a home internet connection to continue training overnight (if you wish to). Whilst not essential, we recommend you have a fundamental understanding of Python or a similar language as we will be using Python on the day.