PREMA Racing partners with Monolith to boost IndyCar performance

As development partner to PREMA Racing, London headquartered Monolith will apply its proprietary AI technology to produce bespoke models for the PREMA Racing engineering team, enabling it to use new, highly efficient channels for unlocking on-track performance.

PREMA Racing’s engineering teams will utilise Monolith to implement customised aerodynamic maps and virtual sensors, with trackside anomaly detection validating data as it’s gathered on the circuit.

Monolith said its suite of machine learning solutions will provide highly time-efficient support for rapidly analysing the challenging variables affecting racecar set-up, from track characteristics and weather conditions to driving style.

MORE FROM ARTIFICIAL INTELLIGENCE

Sam Emeny-Smith, Head of Automotive and Motorsport at Monolith, explained that the company’s AI models ‘plug directly’ into PREMA’s current engineering workflow, providing enhanced data analysis that complements rather than replaces their existing setup methodologies.

He added that racing regulations strictly limit the number of physical sensors teams can install.

“Engineers always want more information about vehicle behaviour,” he said. “Our virtual sensors solve this by extracting additional insights from their existing telemetry streams without requiring any hardware changes or modifications to their established data collection processes.”

Emeny-Smith continued: “We're also exploring opportunities to enhance their broader simulation workflows by improving the quality and representativeness of input parameters, ensuring their models better reflect real-world racing conditions while maintaining established engineering processes.”

Monolith’s models are trained using standard telemetry parameters common to any racing team, such as  ride height, wheel speeds, airspeed, engine performance, suspension travel, and steering inputs that are collected as time series data throughout each session.

Emeny-Smith said: “Before building our models, we've developed analytical tools that use machine learning and advanced statistical methods to identify complex variable relationships and correlations that traditional analysis might miss. This helps us understand which physical characteristics truly drive performance and how they interact with each other, ensuring our models focus on the variables that genuinely move the needle for setup optimisation rather than just the obvious ones.”

For real-time trackside accuracy, Monolith validates it models against actual performance outcomes as they're deployed, creating a continuous feedback loop that confirms they perform reliably under changing track conditions. 

“We can also employ additional tools, such as machine learning anomaly detection, if needed, to ensure data quality during critical sessions,” said Emeny-Smith.

Monolith’s partnership with PREMA Racing is the latest in the company’s top-flight motorsport projects, as the business continues to support the British sports car racing team JOTA Sport.