AI-Driven Self-Diagnosing EV Charging Infrastructure

Interrupted EV charging is a common frustration for private electric vehicle owners—but for fleet operators, it can disrupt entire transport systems. When electric buses fail to charge overnight due to charging station malfunctions, dozens of vehicles may be unavailable the next day, impacting operations, passengers, and costs. Improving the reliability of EV charging infrastructure is therefore critical for large-scale electromobility.

Improving EV Charging Reliability with AI-Based Diagnostics

To tackle charging interruptions at scale, the KI-LOAD research initiative is redefining how reliability is built into electric charging infrastructure. By combining AI-powered monitoring, predictive diagnostics, and intelligent fault detection, KI-LOAD aims to significantly reduce downtime and ensure uninterrupted charging operations. The initiative brings together the applied research expertise of Akkodis Research with leading academic specialists from the Technical University of Munich, creating a powerful bridge between innovation, science, and real-world deployment.

Funded by the Bavarian Ministry of State for Economics, Development and Technology, KI-LOAD is designed to support mission-critical charging environments where reliability is non-negotiable. The project focuses on high-demand use cases such as electric bus and truck depots, where even short interruptions can disrupt operations at scale. By increasing system resilience and enabling proactive maintenance, KI-LOAD helps lay the foundation for scalable, dependable e-mobility infrastructure that supports the transition to zero-emission transport.

Understanding the Root Causes of Charging Interruptions

Charging failures can stem from multiple sources, including unstable vehicle-to-charger communication, hardware defects, and insufficient maintenance practices. These issues are often handled reactively, without structured analysis or long-term prevention strategies.

KI-LOAD introduces a structured, data-driven approach that not only resolves faults faster but also prevents malfunctions before they occur.

Fail-Safe EV Charging for Electric Vehicle Fleets

The project aims to create a fail-safe charging infrastructure that significantly improves vehicle availability for fleet operators. Akkodis Research is working with one of the largest urban transport providers in Northern Germany, which operates hundreds of electric buses across its network.

This real-world deployment ensures that AI-powered diagnostics are validated under operational conditions.

Edge AI and Smart Charging Communication Protocols

KI-LOAD leverages Edge AI to analyze charging data in real time, enabling immediate detection of anomalies. The system uses modern charging communication standards such as OCPP 2.1 and ISO 15118-20 to create a holistic view of charging stations and vehicle battery systems.

Insights are integrated into the central Charge Point Management System (CPMS), enabling automated diagnostics and predictive maintenance scheduling.

From Charging Data to Predictive Maintenance

Akkodis’ EVAcharge charging communication control software plays a pivotal role in the KI-LOAD project by enabling secure, high-quality, and reliable data exchange between electric vehicles and charging stations. Developed and continuously enhanced since 2012, EVAcharge has evolved into a mature, field-proven solution that underpins interoperability across complex charging ecosystems. Today, it supports more than 30% of DC fast-charging points across Europe, making it a critical backbone for scalable and resilient e-mobility infrastructure.

Developed and continuously enhanced since 2012, EVAcharge has evolved into a mature, field-proven solution that underpins interoperability across complex charging ecosystems

The insights and innovations generated through the KI-LOAD initiative will be directly integrated into downstream EVAcharge services, further enhancing their intelligence, reliability, and performance. By embedding AI-driven diagnostics and predictive capabilities into EVAcharge, Akkodis is strengthening its end-to-end EV charging portfolio—helping operators minimize downtime, optimize maintenance, and deliver consistently reliable charging experiences at scale.

Reducing Downtime and Strengthening Electromobility

The project targets a 50% reduction in charging station downtime caused by faulty components and an increase in charging reliability from 95% to at least 97.5%. These improvements are critical for commercial EV fleets, where charging failures lead to operational disruption and increased costs.

By increasing reliability and trust in EV charging infrastructure, KI-LOAD contributes directly to the scalability of electromobility—one of the key technologies in the fight against climate change.

​