Analysing EV Batteries with AI

Francisco Ribeiro Mansilha, a third-year Applied Data Science & AI student, completed a one-semester work placement at INNER, a start-up on the High Tech Campus in Eindhoven developing non-invasive diagnostic technology for electric vehicle batteries. He joined as a computer vision intern, applying AI techniques to analyse battery CT scans.

The work

Francisco developed vision algorithms to detect internal battery structures and anomalies. As he explains, electric vehicles often encounter battery issues after prolonged use, and traditional diagnostics require disassembly. He worked with two other interns on anomaly detection models and helped refine the company’s internal tooling.

What he took from it

The multidisciplinary setting turned out to be the most valuable part:

“At BUas, I usually collaborate with students from similar backgrounds, but INNER brought me into a multidisciplinary environment with people from electrical engineering, mechanical engineering, and applied physics.”

His supervisor, with a background in medical software and computer vision, mentored the team in interpreting CT scans and building robust AI solutions.

Being at a start-up gave him exposure that a larger company would not have — he sat in on co-founder decision-making discussions and produced reports for consulting work analysing EV batteries for external clients.

Challenges

CT scan analysis was initially overwhelming. Francisco worked through it with team communication, research and collaborative problem-solving; the small team made mutual support easy, and the placement ended with a comprehensive report of their findings.

What’s next

Francisco is going on to a second placement at CSEM in Switzerland, again focused on computer vision, before starting his graduation project at BUas. The placement shifted his interests too — alongside computer vision he has developed a curiosity about robotics and electrical components, where “the complexity and innovation involved are really exciting.”