BMW has formulated seven basic principles for the use of AI in the company. These build on the basic requirements formulated by the EU on trustworthy AI and will be continuously developed across all areas of the company.
The seven principles of AI cover:
- Priority of human action and human supervision
- Technical robustness and safety
- Privacy protection and data quality management
- Transparency: Explanability of decisions through AI applications
- Diversity, non-discrimination and fairness
- Social and environmental well-being
The company has been working on this since 2018 when it launched “Project AI” to ensure that the AI technologies are applied in an ethical and efficient manner. This project has now grown into a competence centre at Group level for data analytics and machine learning.
As a hub for knowledge and technology exchange within the BMW Group, the project plays a key role in the digitisation of the company. Among other things, a portfolio tool was developed to create transparency in the application of technologies that make data-driven decisions.
Examples of applications from BMW’s AI portfolio:
AI-supported energy management in the vehicle: There are a large number of electrical consumers in the vehicle, such as the seat heating system, the entertainment system, the air conditioning system and many more. In many cases, the driver is not aware that the use of these consumers also has an impact on CO2 emissions or the range of the vehicle. AI experts from the BMW Group are researching and developing AI-based software for energy management in the vehicle. Based on user behaviour and available information about the route, the system learns to optimally adjust the energy consumption in the vehicle to the driver’s needs and energy efficiency.
Acoustic Analytics: Sensory enhancement in the sensor model for automated driving functions. In connection with the topic of environment detection, the company is researching how AI sensor fusion can be extended to include acoustic signal processing. The integration of auditory perception can bring advantages in the future, especially for urban scenarios.
Linking AI with the control of systems and robots: A controlling AI recently celebrated its premiere at the BMW Group with a smart application in the Steyr plant (Austria). This accelerates logistics processes by ensuring that empty containers avoid unnecessary trips on conveyor belts.
AI in vehicle production: Since 2018, the BMW Group has been using various AI applications in series production. One focus is on automated image recognition processes. Here, the AI evaluates images of a component during ongoing production and compares them with hundreds of other images of the same sequence within milliseconds. In this way, the AI detects deviations from the standard in real time and checks whether, for example, all the intended parts have been installed or mounted in the right place. Flexible AI-based applications are now gradually replacing permanently installed camera portals at the BMW Group. A mobile standard camera is all that is needed to take pictures during production; employees take pictures of the relevant component from different perspectives and mark possible deviations on the pictures. In this way, they create an image database in order to set up a neural network that will later independently evaluate the images.
- The morals of algorithms: a contribution to the ethics of AI systems
- Continental employs Nvidia supercomputer for AI training
- Artificial Intelligence and future directions for ETSI
- Continental develops ethical rules for AI systems
Other articles on eeNews Europe
- RISC-V boom from edge AI says Facebook’s chief AI scientist
- Robots team for autonomous pizza making and delivery
- Graphene electronics pilot line takes shape
- Opinion: Data centre AI drives $30bn AMD-Xilinx deal