With PwC’s AI expertise, Innoptus Solar Team is racing ahead

Launched in 1987, the World Solar Challenge is considered to be the ultimate test for solar car innovation. Belgium first entered a team – from Group T, now part of Leuven University – 20 years ago and in 2019 won the title of world champion for the first time. The team repeated that accomplishment in 2023 and is continually seeking to improve not just the car, but the team’s performance as a result too. And not just because the rules say they must - for each edition of the World Solar Challenge the rules are adjusted to push innovation ever further-; but for the challenge of seeing what they can achieve and how they can maintain their leading position. But they’re not doing it alone. The team of 20 engineering students is backed by 160 different companies - and PwC’s one of them.

PwC began collaborating with Innoptus Solar Team, as the Belgian team is called, in 2023 via sponsorship and we’ve recently extended that to offer our technological expertise too, in the form of artificial intelligence (AI). Explains Kevin Vandeputte, responsible for public and business relations for the team, “this is the first time we’re incorporating AI into the car itself and, as we have limited time and resources to be able to investigate AI ourselves, the expertise PwC brings is essential. It gives us an advantage on the AI front that we can build on for the coming years to help put greater distance between us and other teams, we hope.”

The predominant project we’re working on with Innoptus Solar team is anomaly detection. “The car has many, many parameters that need to be monitored, all the issues that can occur during a race need to be handled, which is manually done at the moment. As the number of parameters keeps growing – there are currently over 200 sensors in the car that need to be tracked - it’s become harder and harder for them to keep up. Automating this will help anomalies stand out so we can act more quickly,” says Thibaut Deliever, Digital Twin Engineer. Kevin Vandeputte adds, “when things go wrong, it can cost you a lot of time in a race and can be the deciding factor if you come in first or second. The monitoring system will help us prepare for these situations as we can see if something will go wrong before it does and whether we already need to change some settings on the car, etc. And even to start thinking about how we can prepare for a future repair while driving, so that we can carry it out quickly once necessary.”

We’re working with the solar team to deploy AI models trained on historical test and race data to efficiently identify anomalies. By tailoring these models to the unique needs of the team, we enable members to capitalise on AI-driven intelligence to continually refine strategy and improve operational performance. We’re also advising on best practices for synchronising data between local and cloud-based InfluxDB systems to ensure seamless and reliable data flows. A systematic labelling methodology will support more nuanced data analysis to allow Innoptus Solar Team to extract deeper insights and well-informed decisions.