More efficient charging and driving: A new approach to autonomous fleet management

Researchers at Sweden’s Chalmers University have developed an optimization method designed to enable autonomous electric taxis to navigate city traffic more efficiently. The approach combines AI predictions with mathematical models and is intended to account for unpredictable factors such as unusual weather conditions.

Image: Waymo

Autonom driving electric taxis not only have the potential to make city traffic more efficient and environmentally friendly but could also reduce costs per trip by up to 60 percent compared to traditional taxis, according to a research report from Stanford University. However, a major challenge in operating autonomous taxis is fleet coordination.

To address this issue, a research team from Chalmers University in Sweden, in collaboration with Volvo Cars, is using a combination of machine learning and classical optimization. To predict passenger demand and available charging stations, the team employs so-called Bayesian Neural Networks (BNNs), which were trained using historical taxi data from San Francisco and Chicago.

The key advantage of this machine-learning method is that BNNs not only calculate pure probabilities but also provide estimates of how uncertain their own predictions are. Even if passenger behavior changes suddenly—such as due to sudden rain—the fleet management remains stable as a result.

The AI does not make autonomous decisions but serves purely as a data source. The actual scheduling of driving and charging is handled by a mathematical optimization model with fixed rules. Since a large city has too many variables, the algorithm breaks down the overall task into smaller sections and combines them into a comprehensive plan. This enables decisions to be made in fractions of a second—even with very large fleets of vehicles.

In addition, the team examined three different types of electric vehicles with varying battery sizes and efficiency levels. The results show how significantly the chosen vehicle technology affects fleet availability. By properly balancing battery size, charging scheduling, and data forecasting, the deployment of autonomous electric taxis in everyday use can be made much more reliable.