We were able to run many DeepRacer events in Buenos Aires in 2023; I thought about writing this article to describe how things were put together. The bottom line is that I wanted to gain experience in physical racing but needed a track. Many racers face the same problem, and I get asked how I did it. Finding a university and building a DeepRacer community around it was the solution for me.
Intro
I have been a racer for some time. I started racing in 2020 when DeepRacer announced the F1 contest. I made it to the finals that year, but there was no physical final as re:Invent was virtual because of the pandemic. That was when I also got my first DeepRacer Evo car as a prize for participating. I also joined the AWS Community Builder program and started talking and blogging about DeepRacer.
The competition in 2021 and 2022 was hard, and I couldn’t make it to the finals again. However, I kept trying and trying. On the other hand, my DeepRacer car was sitting in my Home Office, accumulating dust. I wanted to gain some practice in a physical track, but there are few AWS Summits in LATAM where the track is available to race. I had the money to buy a track myself but I needed to figure out where to place it as I live in a small apartment in the middle of Buenos Aires.
On the other side of the world, Pablo Inchausti (another AWS Community Builder from Buenos Aires) was at AWS reInvent 2022 and got obsessed with DeepRacer when he saw cars racing on the track. Pablo is also a teacher at UADE, so he thought it would be great for the university to give their students some experience with that technology. We have been working as a team since then.
Pablo and I started to talk. We spoke with the university. Daniel Feijó, the director of Computer Science Engineering, quickly saw the potential of this opportunity, and they built the track and bought their own DeepRacer Evo cars. The university also had at least five locations to assemble the track.
Since then, we have organized different races, and the university has always helped host the races in different places and helped with the logistics around those races (transporting the track to different places, assembling the track, providing facilities for events, etc.)
UADE Races (Engineering Projects)
In Computer Science Engineering, students must present a final thesis/project. Three groups of students started working on DeepRacer-related projects:.
- Autonomous Vehicles Applied to Object Detection in Logistics Centers
- Autonomous Vehicles Applied to Inventory Management in Logistics Centers
- Optimizing Machine Learning Models for Autonomous Vehicles
The first few times we built the track was for these students to test their DeepRacer models.
Nerdearla Technology Conference
Nerdearla is one of the most significant technology events in South America. It is a 3-day event and more than 10.000 people attended the conference in person and 27.000 virtually. We gave a virtual workshop before the in-person event so that people could see what DeepRacer was about and race. We had 20 racers competing on the track and the best lap time was 18 seconds. If you would like to know more about this race, you can read more in this blog post.
AWS User Group Buenos Aires Race
We also hosted a race for the local AWS User Group at UADE Belgrano. We assembled the track in a Garden house in an open park across the university’s street. We had about 18 racers, both from the local community and students. The best lap time was 10.2 seconds. Many neighbors stopped by to see what we were doing. After the race in the park, we had a small reception at the university.
UADE Application Development 2 Subject Race
During the last semester, I was offered the opportunity to participate as a teacher in one of the UADE courses that Pablo was running. The subject is about different ways for applications to integrate. Everyone works on the same project: The classroom is divided into groups; each group works on its own microservice, and then it needs to integrate with the other groups. The project was about an autonomous food delivery service, so we used DeepRacer to learn about how ML models work and we ran an internal race as well.
UADE’s own use
UADE has been using the track at science fairs and taking it across different parts of the country to showcase the DeepRacer technology, including high schools.
Final Notes
As you might see in this article, we ran several races for different groups. Hosting all these races was fun, and we are starting to build a DeepRacer community in the region, and that is rewarding. If you are facing the same problem I had: You want to get your hands on a physical track but don’t have the money or the space, reach out to the Universities in your region; one of them might be interested in collaborating with you. It is definitely worth it for both parties.
We are planning a wide variety of activities next year, from doing DeepRacer activities in high schools and universities to running races at Community and Technology events in the region. I’ll be posting about them on my LinkedIn page.
If you are in the region, feel free contact me through my LinkedIn or join the AWS DeepRacer Community on Slack, where you can get in touch and share experiences with me and other DeepRacer community leaders from across the globe.
Extra
