We use Machine Learning to predict charge point failures in advance, allowing operators to hinder downtime and reduce repair costs. The problem is a low stability across the charge point infrastructure network of many European countries, with it being reflected in both consumers and legislation. Predictive maintenance (PdM) enables overview and control of charge point networks. It allows for pre-emptive downtime through carefully scheduled reboots, updates or maintenance. It can help ensure high levels of reliability, an infrastructural foundational need of upmost importance in the transition to electrical vehicles for all providers, operators and nations. Our core goal is to achieve high model performance. By attaining high performance, our model will position Lumina as the preferred predictive maintenance partner in the EV market.