Product reviews, deals and the latest tech news

How the BMW Group has embraced AI to increase sustainability and find useful applications

The automotive business is one of several that is benefiting from AI, which is having a significant influence. Fully autonomous vehicles are one of the most promising applications of AI, but that’s not all it’s doing. Microsoft and Mercedes-Benz, for instance, are cooperating to boost the effectiveness of automobile production.

This week at the AWS re:Invent cloud conference, BMW Group described the influence AI has had on the company and the new applications of AI that will lead to profitable results in the future.

Marco Görgmaier, general manager of data transformation and AI at BMW Group, recently shared that his team has amassed thousands of data assets that may be utilised in future research and AI projects. His team, he claims, has delivered over $1.1 billion in value through more than 800 use cases since 2019. Applications are found in many different areas, including R&D, logistics, sales, quality, and the supply chain.

“Our team’s goal and objective is to accelerate and grow commercial value generation using AI across our value chain,” explained Görgmaier.

BMW’s use of AI aids its drive toward a more sustainable future

Helping to increase sustainability is a new focus area for BMW.

Approximately 70% of the world’s greenhouse gas emissions are produced in metropolitan and peri-urban regions, where 60% of the world’s population resides, as noted by Görgmaier. What BMW is doing currently is trying to help city planners find solutions to challenges that will help cut down on emissions.

BMW is already pitching in, using machine learning algorithms that can foresee how changes to traffic laws would affect vehicle velocity and fuel consumption. In addition, ML models are used to pinpoint areas that still need to be equipped with charging stations for EVs. According to Görgmaier, the shortage of charging stations has a negative effect on sustainability since it discourages people from making the conversion to electric vehicles.

Predicting how parking costs and availability will affect people’s propensity to drive is another area of focus for BMW ML. Commute times and traffic patterns are examples of such patterns that can have an effect on pollution levels.

Amazon SageMaker as a tool for geospatial analysis

According to Görgmaier, geospatial data can help with many of the urban sustainability problems that BMW is working to address. There, BMW is beginning to make advantage of the recently unveiled geographic features of the Amazon SageMaker ML tool package.

BMW hopes to use geospatial ML for forecasting the time when a company with a fleet of cars may make the switch to electric vehicles.

“We set out to train machine learning algorithms to understand connections between engine type and driving characteristics,” he explained. The thinking behind it was that if such a link existed, the model could be trained to anticipate which drivers would be more interested in purchasing an electric vehicle.

BMW had to employ GPS logs and geographical data to build the connections because it was dealing with totally anonymised data at the fleet level.

After its training was complete, the algorithm “was able to forecast how likely it was for certain fleets to switch to EV with accuracy of more than 80%,” as explained by Görgmaier.