Microsoft and Mercedes-Benz tap AI for auto production efficiency

Microsoft and Mercedes-Benz announced a new collaboration today, tapping artificial intelligence (AI), digital twins and data analytics to provide real-time feedback that improves production efficiency, solves supply chain bottlenecks and dynamically allocates resources to prioritize EV and top-end luxury vehicles.  

The new MO360 data platform, an evolution of Mercedes-Benz’ digital production ecosystem, will connect 30 passenger car plants worldwide to Microsoft Cloud. According to a Microsoft press release, the data platform, standardized on Microsoft Azure, provides the flexibility and cloud computing power to run AI at a global scale while addressing cybersecurity and compliance standards across regions — it is already available to teams in Europe, the Middle East and Africa and will be deployed in the United States and China. 

“We’re making the transition to electromobility even faster and more efficient,” Jörg Burzer, member of the board of management of Mercedes-Benz Group AG, production and supply chain management, told VentureBeat. “The data platform provides us with AI technologies that make our production more efficient, our supply chain more resilient and also pushes sustainability to the next level.” 

Microsoft’s real-time digital feedback loop

With the new MO360 data platform, the Stuttgart-based automaker can create a virtual replica, or digital twin, of its vehicle manufacturing process, combining insights from assembly, production planning, shop floor logistics, supply chain and quality management. The virtual simulation and optimization of processes, before running them on the shop floor, helps to accelerate operational efficiency and unlock energy savings. For example, managers can optimize operational patterns to reduce CO2 emissions in production.  

“What’s different is the real-time digital feedback loop that’s being created by the industrial metaverse,” Judson Althoff, Microsoft’s EVP and chief commercial officer, told VentureBeat. 

Most AI projects to date, even the very successful ones, tend to be sort of an offline analytic exercise, he explained. “In other words, let’s have some machine learning, operate against a data set, leverage some cognitive services to study a problem, then let’s learn from it and implement something new or different based on what we’ve learned,” he said. “In this scenario, you actually have the digital twin teaching the people and then the people in turn teaching the model for continuous learning in real time.” 

Virtual replica of Mercedes manufacturing process

Althoff added that he is coming off a “bit of an emotional high” after touring Factory 56, a 220,000-square-meter, fully-digitized production facility at the Mercedes-Benz plant in Sindelfingen, Germany, which is enabled by the MO360 data-driven AI platform. 

“MO360 is the brains behind the facility … this is really one of the best marriages of technology business outcomes and societal outcomes that I’ve ever seen,” he said. “In every step along the way, there’s been a super-pragmatic approach towards making sure the technology is actually in service of the outcomes — both from the business goals of increased production at lower cost, as well as the sustainability goals that Mercedes-Benz has not just in the vehicles themselves, but in the manufacturing process.” 

Althoff offered the example of line workers responsible for installing car seats watching cars go by on the shop floor. “They are really great at managing the quality of upholstery and, you know, just the overall ergonomics of the car cockpit — they don’t want to spend their lives chasing down parts and wondering what car might be coming next in the factory,” he said. 

The data model that underpins the entire factory in MO360, he explained, uses digital replicas to model what parts are required for each car coming down the line. A system of robots delivers the parts to the highly-skilled workers and the workers assemble the cars one by one, no matter what sequence of cars might be happening — whether it’s a Maybach or S-Class. 

Microsoft democratizes the digital experience

In addition, MO360 brings humans and AI together, democratizing the digital experience, Althoff emphasized. 

“This is not about some ivory tower replacement of the frontline worker, but rather an empowering set of assets that the people on the frontlines can use,” he said. “We’re using digital replicas to simulate the entire end-to-end process, so each worker understands now how to get better at their job based on the learnings of the past.” 

For example, he went on, one of the big learnings has been about the torque settings of every single bolt on the car, which is stored, recorded and audited in the cloud. “So it’s sort of a lifetime association with the vehicle itself for safety records and things like that,” he explained. 

“It’s an end-to-end process view that has never been able to be simulated before … that is really AI empowerment of every employee from the plant manager and engineers down to the people on the lines, all working in synchronicity with the AI model.”

Automotive production of the future

“I’m convinced that AI, and especially what Microsoft is doing together with us, is the only way you can bring this transformation towards electromobility,” said Burzer. “Now we have a unique opportunity to incorporate this into the automotive production of the future, which is basically already becoming reality today, at least in Factory 56.” 

In the future, he added, he also hopes to continue to bring line workers together with AI even more deeply. “I’m personally touched by that because you can really see our team members on the line at the shop floor using our MO360 applications now every day, and we need our team members to build these first-class cars,” he said. “We can basically put that all together now with the help of Microsoft … this is one of the topics I would like to push even further.”