Accelerating Autonomous Racing
The Indy Autonomous Challenge (IAC) is a global competition in which university teams develop autonomous driving software for race cars that are capable of safely operating at speeds over 305 km/h (190 mph) – without human drivers. The goal: to advance autonomous driving technologies under extreme conditions.
As a proud technology partner, dSPACE provides every IAC team with cutting-edge simulation tools and high-performance hardware. From cloud-based virtual test environments and real-time sensor simulation to onboard computing, dSPACE solutions help the teams design, test, and deploy AI drivers with confidence.
Which dSPACE products are used in the IAC?
- SIMPHERA: a simulation environment that enables realistic testing and further development of AI driver software. It includes simulation models, test automation, and test analysis. The Cloud Racing Services, developed specifically for the IAC, offer a virtual test environment hosted in the AWS cloud, allowing flexible resource allocation and parallel execution of test scenarios.
- AURELION: a sensor-realistic simulation software with which the teams can realistically test perception algorithms on virtual racetracks. The AI driver software receives simulated sensor data from cameras, radar, and lidar, processing the data just as it would in a real racing car.
- AUTERA AutoBox: a robust in-vehicle platform that serves as the central computer in all IAC race cars. It reads and processes data from lidar and radar sensors, cameras, and buses and networks in the vehicle. The unique combination of high computing power, ruggedization, and high-speed logging capabilities in a compact form makes the AUTERA AutoBox especially valuable for the IAC.
"In the first year of our partnership with the IAC, we were able to experience the competence and enthusiasm with which the teams use the pioneering technologies to further develop the cars. Now that the AUTERA AutoBox has established itself as the brain of the IAC race car, we are now taking the next step and providing the up-and-coming engineers with a virtual racetrack. With the software-in-the-loop solutions, we equip them with powerful tools that enable them to gain insights early on and that make the difference in the end to win the IAC race series."
Dr. Carsten Hoff
CEO, dSPACE
"dSPACE has been a cornerstone partner for the IAC, playing a pivotal role in our world’s fastest autonomous racecars. Elevating our collaboration, dSPACE will provide a customized version of their SIMPHERA SIL product, a best-of-class high fidelity simulation tool that will offer our university teams a true digital twin of the IAC AV-24 racecars and track environments. This partnership will enable true SIM to REAL transfer learning that will unlock the full potential of our AI drivers to reach the pinnacle of safe high-speed automation."
Paul Mitchell
CEO, President, IAC
What the IAC Teams Say About the dSPACE Simulation Solution
The simulator is absolutely vital for safely testing our software on new tracks, giving us immediate feedback on how the car behaves in unfamiliar layouts. Moreover, when developing new planning algorithms, having a simulator that is able to replicate both vehicle dynamics and interaction with race control has proven fundamentally important.
Caltech Racer uses the dSPACE simulation solution to run virtual races as part of our verification and validation pipelines. In particular, we use dSPACE products to almost completely eliminate the track-specific integration while on-site, significantly improving our test slot utilization as we know our race lines and configurations will work before we even get to the track.
The dSPACE simulation environment has been pivotal for our team. It allows us to rigorously test and validate algorithms before hitting the track, which significantly bridges the sim-to-real gap and saves us invaluable time. The responsive support from the dSPACE team has also been instrumental in accelerating our development.
The dSPACE simulation environment is an extremely useful tool for developing and validating our autonomous driving software. For teams like ours based in Asia, where access to the track is limited, it allows us to thoroughly test and refine our systems ahead of rare on-track opportunities. The realistic simulation of both the environment and the vehicle model also gives us great confidence in the results.
