Advancing EV Battery Optimization Through Decentralized AI

Today’s electric vehicles rely heavily on Battery Management Systems to estimate parameters such as remaining range, charging capability, and battery health. However, most existing systems are still based on static laboratory models applied generically across entire vehicle fleets. In practice, batteries age differently depending on driving behavior, charging patterns, climate conditions, and operational history. 

This often leads to significant inaccuracies in State-of-Charge (SoC) and State-of-Health (SoH) estimation, resulting in reduced usable performance, inefficient charging strategies, unnecessary battery oversizing, and conservative safety margins. 

While cloud-based analytics can partially improve diagnostics, they also introduce challenges related to latency, data transfer costs, connectivity dependence, and strict data privacy regulations such as GDPR. In many cases, only limited data can realistically be transferred from vehicles, restricting the ability to continuously adapt battery models to real-world conditions. 

Moving AI Directly to the Battery System 

To address these challenges, Cognivity AI Sweden AB, together with a California-based luxury EV manufacturer, has through Future Mobility secured funding from Vinnova and the Swedish Energy Agency to investigate how decentralized artificial intelligence can be integrated directly into existing battery systems. 

The project explores how AI models previously validated in stationary energy storage applications can be adapted to the demanding and dynamic conditions of the automotive sector. Instead of relying solely on centralized cloud processing, the architecture allows intelligence to operate locally on the battery system itself. 

This decentralized approach enables batteries to continuously adapt to their individual usage conditions while sharing learned insights between vehicles without transferring raw user data. The result is a scalable framework for fleet-level learning with significantly improved data privacy and cybersecurity

The project specifically focuses on three main areas: 

• Decentralized Architecture and Data Security 
  Investigating how AI can operate directly on the battery system while maintaining GDPR compliance through privacy-preserving learning approaches.
• Improved Diagnostic Accuracy 
  Evaluating how AI-driven models can improve estimation accuracy for State-of-Charge (SoC) and State-of-Health (SoH) under real-world operating conditions using laboratory and engineering fleet data.
• Cross-Platform Learning and Integration 
  Exploring how multi-layer machine learning architectures can be integrated using standard measurements such as pack current, cell voltages, and temperatures, without requiring additional hardware or sensor costs. 

By enabling batteries to learn from operational behavior over time, the project aims to create more adaptive battery systems capable of improving charging performance, extending battery lifetime, and increasing operational reliability. 

Bridging Swedish AI Research and Californian Automotive Innovation 

“By collaborating across borders with a California-based luxury EV manufacturer, we are able to combine Sweden’s strengths in AI and predictive analytics with the fast-moving automotive innovation ecosystem in California. One of the biggest advantages is the ability to work closely with real engineering challenges from the vehicle side, access relevant data quickly, and validate ideas in an environment where new technology can move from concept to testing very rapidly. 

This type of collaboration creates a much shorter path between research and real-world deployment. It allows us to refine our models using practical use-cases while also solving challenges around secure international data collaboration and privacy-preserving AI. The result is not only better battery diagnostics and performance, but also a stronger foundation for the next generation of intelligent and adaptive battery systems.” — DR. Christian Fleischer, CEO, Cognivity AI. 

Supporting the Future of Sustainable Mobility 

By combining decentralized AI, predictive battery analytics, and real-world automotive engineering, the collaboration aims to establish a validated foundation for future AI-driven battery management systems in electric vehicles. 

The long-term objective is to support more efficient, reliable, and sustainable transport systems through improved battery utilization, reduced degradation, and more intelligent energy management. 

Do you want to know more?