CoreWeave's AI Tool Enhances JOTA's Race Car Testing Efficiency
With CoreWeave's AI recommendation tool, JOTA's engineers can now optimize race car setups in record time, transforming testing into a data-driven, efficient process. Discover how this technology is revolutionizing motorsport engineering.
Key Facts
- CoreWeave's AI tool improved JOTA's testing efficiency, maximizing learning from limited rig time.
- JOTA's engineers reduced setup testing from millions to targeted configurations, enhancing performance.
- Active learning in AI revealed setup uncertainties, giving JOTA a competitive edge in decision-making.
- Collaboration with CoreWeave cut time-to-insight, potentially increasing JOTA's financial returns on tests.
- Trackside AI application extends benefits, indicating a strategic shift towards real-time performance optimization.
Summary
Summary
JOTA, a motorsports engineering team, faced challenges in efficiently testing race car setups due to limited time on their suspension rig. CoreWeave provided an AI-powered recommendation tool that analyzed existing test data to suggest optimal setups for further testing. This collaboration significantly enhanced JOTA's learning process, allowing them to maximize the value of each test run.
Background
JOTA is a motorsports engineering team that has worked with various racing organizations, including Hertz Team JOTA and Cadillac Hertz Team JOTA since 2025. Before implementing the AI solution, JOTA relied on traditional methods to test car setups on a seven-post rig, which simulates track conditions. The engineers faced constraints in time and data analysis, limiting their ability to explore the vast number of potential configurations.
Challenge
The primary challenge was the inability to efficiently determine which setups to test next. With millions of possible configurations and only a limited number of runs available, JOTA's engineers struggled to extract actionable insights from the data generated during testing. Each run took about seven minutes, leaving little time for thorough analysis and decision-making.
Solution
CoreWeave's Physical AI Field Engineering team developed a setup-recommendation tool tailored to JOTA's needs. This tool utilized existing test data to identify setups that would yield the most learning opportunities. The collaboration involved defining the problem together and building the deployment on-site, allowing JOTA's engineers to maintain control over the configurations tested while benefiting from AI-driven insights.
Results
The implementation of the AI recommendation tool allowed JOTA to significantly enhance their testing efficiency. By focusing on setups that maximized information gain, JOTA was able to make better-informed decisions on which configurations to test next. The tool improved the learning process during testing, although specific quantitative outcomes were not provided in the source material.
Key Insights
- Collaboration between domain experts and AI engineers can lead to more effective solutions tailored to specific industry needs.
- Active learning techniques can optimize testing processes by identifying setups that provide the most valuable insights.
- Understanding the customer's domain is crucial for developing tools that engineers will effectively use.
Customer Testimonial
“Sometimes we know AI could solve a problem for us, we just don't know how to get there. That's exactly what working with the team gives us, help figuring out the approach, and help getting it into use.” — Tomoki Takahashi, Technical Director, JOTA Group
Entities Mentioned
Companies
Products
Technologies
People
Key Concepts
Definitions
- active learning
- A machine learning approach where the model actively queries for the most informative data points to learn from, improving its performance over time.
- multi-objective performance index
- A metric that evaluates multiple performance criteria simultaneously, allowing for a more nuanced assessment of setups.
- setup-recommendation tool
- An AI-powered tool that analyzes existing data to suggest optimal configurations for testing, enhancing learning efficiency.
- rig testing
- A method of simulating real-world conditions for race cars using specialized equipment to evaluate performance without being on the track.
- trackside application
- An AI tool used during on-track sessions to provide real-time recommendations for adjustments based on performance data.
Use Cases
- →Optimizing race car setups before track testing
- →Recommending adjustments during on-track running
- →Enhancing learning from limited test runs
- →Improving decision-making in high-dimensional datasets
- →Collaborative tool development between engineers and AI experts
- →Maximizing information gain from test configurations
Frequently Asked Questions
How does the setup-recommendation tool work?
The tool analyzes past test data to identify setups that are likely to yield the most valuable insights. It uses active learning to continuously improve its recommendations based on new data.
What are the benefits of using AI in motorsports?
AI helps teams optimize their testing processes by providing data-driven recommendations, which can lead to better performance outcomes and more efficient use of limited testing time.
Can the tool adapt to different racing conditions?
Yes, the tool is designed to learn from various test results and can adjust its recommendations based on the specific conditions and performance metrics of each race.
What role does collaboration play in developing AI tools?
Collaboration ensures that the AI tools are tailored to the specific needs of motorsports engineers, leveraging their expertise to create solutions that are practical and effective in real-world scenarios.
How does CoreWeave's team differ from traditional data science teams?
CoreWeave's team combines expertise in motorsport engineering with AI, allowing them to understand the unique challenges of the domain and create tools that are immediately applicable without extensive domain learning.