The Angevin-based startup, which has very strong growth ambitions, has already proven the viability of its techniques for evaluating the health and lifespan of lithium batteries to be reconditioned. It is now a question of industrializing his project. In this, AI will be crucial, considers its leader.
AI is a necessity for VoltR.
At the same time, it is a question of estimating the lifespan of the battery compared to a performance staircase, making it possible to define its role in its second life.
On average, eight out of ten batteries are not physically damaged and they have an SOH of around 80%.
“VoltR can accommodate any type of lithium batteries in an incoming flow. In outflow, we cannot recondition batteries for the highest levels of performance: those corresponding to the needs of cars, laptops, smartphones,” notes Alban Regnier. “On the other hand, a laptop battery with an SOH of 80% is equivalent to 120% of the performance of a new bicycle battery.”
[...]
The multitude of variables that need to be taken into account when predicting battery life can make the data science exercise energy-intensive and expensive. “We try to be as reasonable as possible when developing a model. In this case, given the complexity of prediction and the need to preserve the lifespan of cells, we use neural network models,” explains Inès Jorge.
The results of the models are compared with real data.
