2024 Recap
2024 was one of UQ Racing’s most successful years, yielding two new competitive cars and excellent competition results. This year, we focused on developing a brand-new Autonomous Vehicle (AV) platform and refining our competitive Electric Vehicle (EV) concept, pushing its performance even further. With a season dedicated to rigorous testing, validation, and execution, our efforts paid off in a big way at competition.
Autonomous Vehicle (AV24) – Dog:
For the past few years, our AV program has been built on the foundation of our most successful EV ever, Chicken, which was retrofitted with autonomous systems like steering, EBS, and an advanced perception package including LiDAR and IR cameras. While this approach worked, it also meant our AV was inherently a step behind our EVs in terms of technology. That changes in 2024.
This year, we introduced Dog, a major redesign that brings our AV up to date with the technology we've developed in our EVs over the last couple of years. One of the biggest changes is in the powertrain and accumulator—our AV now shares the same segment and BMS architecture as our EVs, making our designs more streamlined and integrated. The move to hub motors is another big shift, with dual hub motors at the rear now controlled by our MoTeC M150 ECU. Beyond that, we've made significant improvements to actuation systems, including a safer EBS solution, refined steering actuation, and a revamped service brake design.
While the mechanical and electrical architecture has seen an overhaul, the software side has also undergone serious improvements. Our perception neural network has been made over eight times more efficient, now running at an average of 10-15Hz, a huge step up from previous iterations. The perception package has also been migrated from Python to C++ for faster data processing and real-time evaluations. Path planning has seen similar advancements, with new control and monitoring systems for path tracking and error correction.
Finally, the team has invested in a brand-new onboard computing platform featuring the Jetson Orin NX chip, providing significantly more processing power for our AV's decision-making systems. The end goal for 2024 was clear: take the solid groundwork from previous years and turn our AV into a fully functional, reliable, and competition-ready platform.
Electric Vehicle (EV24) – Frog:
Not much needs to be said about EV24—because we knew the concept potential was there. Instead of reinventing the wheel, we focused on refining and executing our EV23 concept to its fullest potential. Frog is an evolution of our rear-wheel-drive AMK-powered car with 20” Goodyear tires, but with major improvements in the accumulator. We’ve made a significant jump from a 400V, 4.1kWh pack to a 600V, 6.2kWh design, giving us more energy to push harder in endurance without sacrificing efficiency.
The biggest factor in realising the full potential of this concept was maximising cornering performance. This is where testing became our highest priority—validating tire performance, refining vehicle setup, and ensuring our drivers were trained to extract the most out of the car. From the start of the year, we ran extensive testing sessions, characterising our tires to ensure they reached optimal operating temperatures, performing aerodynamic validation through CFD testing, and running endurance simulations to fine-tune our energy-limiting strategies.
Speaking of aerodynamics, our approach this year was points-first, prioritising real world performances over simulated results. Every design decision was made with competition performance in mind, and we leaned heavily into manufacturability to shed unnecessary weight. The result? Over 75kg of downforce at 60kph, with an aero bias of 46% rearward, significantly reducing cornering sensitivity and ride height dependency. Our on-car aero performance was validated within 7% of our simulations. Additionally, minimum ride height was increased by 15mm, and we shaved 5kg off the overall car weight.
Ultimately, EV24 was about refining and executing a strong concept rather than overhauling it. The goal was to validate our conceptual decisions from last year, implement complex control systems that needed real-world testing, and extract every ounce of performance possible through iteration and refinement.
