How artificial intelligence can help optimize electric vehicles

Electric vehicles (EVs)—from cars and motorbikes to commercial vans and trucks—are no longer the alternative choice. In 2023, almost one in five vehicles sold globally were electric. And by 2035, EV volumes are forecast to top 71.6 million (up from 14.2 million in 2023).

The choice is abundant: every major automaker has at least one electric vehicle in its range. And looming emissions targets, such as the EU’s policy requiring all cars and vans to be zero emissions by 2035, means electric vehicles will become the de facto choice.

However, before we get there, several hurdles are hindering the uptake of EVs, from range anxiety and battery performance, to charging infrastructure and maintenance.

Is artificial intelligence (AI) the answer to unlocking the full potential of EVs?

In perfect battery health

Just like your mobile phone, EV batteries degrade over time. According to research, they lose 2.3% of their capacity per year, potentially reducing their range and reliability. However, emerging AI-powered battery management systems can monitor a vehicle’s battery health in real time, analysing temperature, voltage and current and adjusting these parameters to protect the battery to help prevent thermal runaway, which can cause lithium-ion batteries to overheat and pose safety risks.

With machine learning algorithms, AI can also help to predict when a battery is getting close to a critical threshold, allowing drivers to proactively plan maintenance or a replacement. A recent study found that AI-powered battery management systems can potentially extend EV range by up to 10% through adaptive charging strategies.

In the future, AI could even change the very fabric of batteries. “AI has the potential to supercharge the discovery of complex battery materials and processes, enabling faster charging, higher energy density and improved sustainability,” said Murtaza Zohair, an IBM research scientist in the Almaden lab.

“Low-energy consumption AI chips will help reduce the need for large batteries,” added Noriko Suzuki, from IBM’s Institute for Business Value.

On the charge

Europe, the United States and the UK need six times more public chargers than are currently available by 2030 to meet the growing demand, reports McKinsey. AI goes beyond helping drivers find a charger—it can assist local governments and cities plan the infrastructure.

Because AI can anaylze huge datasets, from traffic patterns to grid capacity and EV adoption rates, it can predict where demand will be highest, allowing installers to prioritize their placement.

Once installed, AI can monitor the charging stations to detect potential problems such as slower charging rates and unusual heat levels, to pre-empt problems and schedule maintenance before drivers experience problems. These smart chargers can even manage capacity by dynamically changing pricing based on demand.

Problem predictor  
 

It isn’t only charging stations that benefit from predictive maintenance. While EVs have fewer moving parts than their petrol or diesel equivalents, diagnosing problems early is the key to avoiding trouble.   

If you've ever watched Formula One, you might have seen the engineers remotely monitoring the cars on track. AI is able to do the same thing, working in the background and analyzing vast amounts of data to help predict mechanical issues before they arise. It can detect motor wear, software faults and, as we mentioned above, battery issues before they become bigger problems.

Autonomous for the people

AI and machine learning is already being used in advanced driver assistance systems (ADAS), taking the data from sensors and cameras to navigate and make decisions. Algorithms can learn and adapt to different driving conditions and behaviours to improve safety, and bridge the gap in communication between your vehicle, other vehicles and even the infrastructure around you.

For example, HERE’s ADAS technology can warn drivers about unforeseen hazards ahead. And HERE, Porsche and Vodafone recently collaborated on a proof of concept using AI to help detect vulnerable road users.

EV does it
 

Getting from A to B in an EV requires a different approach to route planning. If you’re likely to exceed your range, you’ll need to strategically plan your charging stops. It’s not only distance that can affect range, with terrain, traffic and weather all playing a role in how far you can go.

Researchers from the Arab Academy for Science in Giza, Egypt, discovered that wind speed and direction could save up to 49% of the battery’s capacity during a 50km journey, and opting for a road with fewer inclines could save 46%.

AI-powered navigation goes beyond planning the shortest or quickest route, with dynamic, real-time navigation dynamically routing and rerouting to keep you on the most efficient path.

For instance, HERE AI Assistant provides intelligent EV routing, taking into account factors that can affect range for the most energy-efficient travel routes. It also uses natural language, so you could, for example, say to your navigation system: “calculate a route to Las Vegas and find a place to charge 50 miles before arrival with a coffee shop”.

AI is making electric vehicles easier to live with. It helps keep batteries in good shape, finds nearby charging stations and plans efficient routes. It can even predict when maintenance is needed and predict future charger availability.

In short, AI is helping EVs run more smoothly and fit more easily into everyday life.