AWS Public Sector Blog
UNSW students build an all-electric race car with AWS
Throughout 2023, the students from Redback Racing at the University of New South Wales (UNSW) wove together their many disciplines of engineering prowess to create their latest cars: RB23 and RB21-D. After developing and going live with their real-time telemetry system on Amazon Web Services (AWS), the team has placed as the highest-ranking Australian squad in the electric vehicle (EV) division of the Australasia Formula SAE competition, placing second overall.
What goes into an electric race car?
Redback Racing is a student-led project at UNSW under the School of Mechanical and Manufacturing Engineering. Established in 2000, each year, the team of more than 120 students design, build, and race a formula student race car in the annual Formula SAE-Australasia competition against various domestic and international universities. After completing their first all-electric car, RB21-E, in 2022, the team began 2023 more determined than ever.
The engineering of RB23 began with a streamlined design philosophy of “keep it simple; do it right,” evolving on the design of RB21-E. The various teams have their own portfolios for optimizing the vehicles, such as aerodynamic features, electric motors, batteries, suspension, and more. The engineering teams depend on real-time data from various components to analyze performance, allowing the teams to make informed decisions on areas of prioritization and improvement. This is facilitated by the software and architecture created by the Data Acquisition (DAQ) team, who develop and deploy cloud technology to make real-time telemetry possible. The seamless integration of cloud-based solutions has allowed the various teams to collaborate and develop more freely, creating an even better vehicle.
Redback Racing’s RB-23 electric race car in action.
Race-ready architecture in the AWS Cloud
The Redback Racing team required a cloud environment with the capability to integrate the various requirements between their many teams, with the goal of creating the fastest and most reliable race car possible. The AWS Cloud was their choice in building out each component required.
Zero to a hundred as fast as possible
Creating a real-time telemetry system to keep up with a race car posed two main challenges to the team: How do we minimize latency, and how do we scale the solution? To solve these issues, the team utilized AWS Fargate with Amazon Elastic Container Service (Amazon ECS) to create their custom telemetry solution. Fargate’s serverless compute engine allowed the team to seamlessly scale their telemetry system to meet race demand and scale in to minimize costs outside of racing. This was accomplished without the team having to provision or manage underlying virtual machines themselves. The team minimized latency in the architecture by streaming from the car in real-time to an Application Load Balancer in front of their ECS cluster, while logging to Amazon Simple Storage Service (Amazon S3) and Amazon CloudWatch. This allowed the team to retroactively analyze the car’s data in addition to real-time dashboards.
Fostering a collaboration station
In addition to the use of the AWS Cloud for their data and processing needs, Redback Racing uses AWS Amplify as the hosting solution for both its team website and the dashboard in front of its telemetry architecture. Amplify simplifies the process of connecting repositories to hosted environments and deployments, allowing the team to quickly test new features and deployments. The team also utilizes Amazon Route 53 for their domain name management and Amazon Cognito to manage the authentication and access to their telemetry dashboard through the public internet.
Redback also relies on its cloud environment to collaborate on design between its various teams, using Amazon Elastic Compute Cloud (Amazon EC2) to host the vital servers behind engineering data access, such as CAD documents, engineering change documents, and bills of materials. Amazon EC2 allows the team to have these servers consistently available, allowing collaboration of various portfolios remotely.
Putting an architecture to the test
In December 2023, Redback Racing finished second overall in the EV Formula SAE-Australasia competition at Calder Park Raceway, Victoria. Ensuring the smooth running of their car on track not only represented a year of hard work by the students, but also the paramount safety of the driver within the car. The stress of a high-speed race on electrical accumulators, batteries, and motors requires a reliable real-time telemetry system to monitor them. The team’s telemetry system on the AWS Cloud allowed the team’s on-track race engineers to evaluate the car in real-time, as well as display live warnings to the driver.
Brandon Nanthathammiko and Shady Alwidyan, Redback Racing’s Data Acquisition leads, stated, “AWS played a vital role in allowing us to secure a place on the podium. During our endurance event, our trackside engineers were able to view and evaluate real-time sensor values reported from RB23, allowing them to make informed decisions as to how RB23 should be driven. Furthermore, it gave them better peace of mind knowing how the car was performing in real-time. Evidently, in the latter half, the throttle demand was reduced to slow down the rate at which the accumulator cell temperatures were rising.”
A Redback Racing Data Acquisition team member tests telemetry at a track event.
Electric race car solutions beyond the horizon
Redback Racing has created a reliable and scalable solution through AWS to not only support their data acquisition team but also their many other portfolios. The solution encompasses real-time data collection, monitoring, logging, analysis, infrastructure hosting, as well as team collaboration tools. The serverless technologies they used empower Redback Racing to focus on innovation within the team rather than maintaining the infrastructure required to function.
Kalindu Dahanayake, Redback Racing’s Team Lead, stated “I am incredibly thankful for the team last year, especially to Leo Cao, Soura Mandal, and Alvin Cherk, for migrating and building the incredible cloud computing architecture we have used as the foundation for 2024’s accomplishments. It has allowed us to use telemetry on all our trackdays in 2024 and was instrumental to our success last year, I expect it to be crucial for us at the competition this year as well.”
Since their foundational AWS developments in 2023, the team have made numerous improvements throughout 2024. The two cars raced in 2023, RB23 and RB21-D, have been developed into a single hybrid vehicle that can be human-driven or autonomous – RB24. On-car CANBUS processing has been offloaded into their AWS environment using AWS API Gateway, AWS Lambda, Amazon DynamoDB, Amazon S3, and Amazon ECS using AWS Fargate. This has decreased latency, improved accuracy and reliability, as well as provided ease of data transformation and analysis. The team have also built out an automated simulation pipeline for their autonomous driving system, utilising API Gateway, DynamoDB, and Lambda.
Redback Racing intend to further build out their capabilities within AWS through 2025, after the next competition in December 2024. This includes live on-car video streaming utilising Amazon Kinesis, online car set-up tracking synchronised with generated logs, automated end-to-end testing from car to cloud, auto-optimisisation of parameters for their autonomous driving system, utilising GPU compute on AWS to test autonomous perception, and extending existing telemetry capability with live mapping data.
Watch this video to learn more about the RB23 car, and watch this video to see an onboard run from the competition.