Greener Fleet Management: AI is the Key to Greater Sustainability

The number of motor vehicles in Germany reached a new high of 60.13 million in 2023. While this is good news for car manufacturers, it is only partially positive for the environment since electric vehicles still represent the minority. The abrupt end of electric vehicle subsidies for private individuals could also significantly reduce further demand. However, electrification remains one of the most important pillars on the path to greater sustainability in transportation.

Against the backdrop of these developments, it remains important—especially for fleet managers—to consider other components such as driving behavior to promote more sustainability. After all, fleet managers can effectively address this issue because data from a large number of vehicles is available—and that's what ultimately matters.

Energy efficiency has various components

The calculation is simple: The more fuel a vehicle consumes, the more greenhouse gases it emits. Therefore, the goal should be to achieve the highest possible fuel efficiency. However, many factors play a role in this. The driving route is certainly obvious: It should be as direct as possible and over smooth roads to the destination, avoiding traffic jams and other obstructions like construction sites and inclines to not unnecessarily increase fuel consumption. Driving behavior also plays a role, as hard braking, late shifting, frequent idling, or fast driving are not energy-efficient. On the contrary, aggressive driving can increase fuel consumption by 20 to 30 percent. Additionally, the tires should be properly inflated because fuel efficiency decreases by about 0.2 percent for every bar under the recommended tire pressure. And the load should be as light as possible and evenly distributed in the vehicle.

In practice, however, fleet managers face the question of how to measure this data and derive actionable insights from it. The solution is telematics platforms that collect this data and AI applications that analyze and provide it. The latter can also be done manually to a limited extent, but it involves considerable effort. Generative AI, based on Large Language Models (LLMs), can analyze huge amounts of data, recognize patterns in it, and derive recommendations from it.

AI delivers analyses and insights

For example, Geotab has launched a test environment called Project G and opened it to its customers, integrating generative AI models into its own platform. It works similarly to what is now familiar from ChatGPT: Fleet managers enter a question into a chat interface and receive an answer within a short time, sourced from the company's specific data as well as billions of anonymized data points from Geotab's databases.

For example, fleet managers can quickly ask which of their vehicles or vehicle groups have the longest idle times and how to reduce them. Based on the collected data, the AI can respond within seconds with various tips. Such insights alone are helpful in increasing operational efficiency or reducing costs. In terms of energy efficiency, they can have data such as employee driving behavior, vehicle performance, and condition analyzed. If the AI concludes that the majority of drivers are too aggressive or at least not energy-efficient in traffic, training on this topic with specific instructions for employees on how to behave may be needed.

Additionally, AI can also monitor the condition of the vehicles and make maintenance predictions. This allows fleet managers to increase the longevity of their vehicles, which is naturally in line with sustainability. At the same time, maintenance costs or even the fuel consumption of older and new vehicles can be compared to determine whether the use of existing vehicles is still cost-effective, or if they should be replaced for environmental reasons.