New AI tool predicts if your electric vehicle can complete planned trips

Engineers at the University of California, Riverside have developed State of Mission (SOM), an AI diagnostic tool that predicts whether an electric vehicle can complete a specific trip based on real-world conditions rather than just showing battery percentage. The system combines machine learning with physics to factor in elevation, traffic, temperature, and driving style, addressing a critical gap in current EV battery management that often leaves drivers uncertain about their actual range.

How it works: SOM replaces traditional battery gauges with mission-specific predictions by blending AI adaptability with electrochemical reality.
• The hybrid model “learns” from how batteries behave over time—how they charge, discharge, and heat up—but stays grounded in physical reality so it can handle surprises like sudden cold snaps or steep climbs.
• Instead of just showing how full the battery is, SOM tells drivers whether their EV can safely and reliably complete a planned journey under current conditions.
• “It’s a mission-aware measure that combines data and physics to predict whether the battery can complete a planned task under real-world conditions,” said Mihri Ozkan, a UCR engineering professor who helped develop the system.

Performance improvements: Testing with NASA and Oxford University datasets showed significant accuracy gains over conventional diagnostic tools.
• SOM reduced prediction errors by 0.018 volts for voltage, 1.37°C for temperature, and 2.42% for state of charge compared to existing systems.
• The team used real-world battery performance data including charge and discharge cycles, temperature shifts, voltage data, and long-term trends.

Why current systems fall short: Today’s battery management relies on either rigid physics equations or opaque AI models, creating uncertainty for drivers.
• “By combining them, we get the best of both worlds: a model that learns flexibly from data but always stays grounded in physical reality,” said Cengiz Ozkan, UCR engineering professor and co-lead researcher.
• Your EV might show 40% charge left, but that doesn’t always mean you’ll make it over that mountain pass with the heater blasting at 65 mph.

The big picture: SOM transforms abstract battery data into actionable travel decisions across multiple applications beyond just EVs.
• “It transforms abstract battery data into actionable decisions, improving safety, reliability, and planning for vehicles, drones, and any application where energy must be matched to a real-world task,” Mihri Ozkan said.
• The system could work with emerging battery chemistries like sodium-ion, solid-state, and flow batteries.

What’s next: The technology faces computational challenges but shows promise for widespread adoption across energy storage applications.
• SOM requires more computing power than typical lightweight EV battery systems can currently handle, though the UCR team is confident optimization will enable integration.
• “The same hybrid approach can improve reliability, safety, and efficiency across a wide range of technologies from cars and drones to home battery systems and even space missions,” said Cengiz Ozkan.
• Details of the research have been published in the journal iScience.