Reducing the time it takes to simulate production flow from days to the click of a button

Democratizing the digital twin

The history of Michelin has been inseparably linked with mobility and innovation for over 130 years. To this day, the company, headquartered in Clermont-Ferrand, France, influences the development of mobility worldwide. With this experience, Michelin addresses current challenges and ventures into new, life-changing business fields emerging from new technology and sustainability.

Using the digital twin to improve efficiency

Global competition and shorter time frames put a premium on being efficient from the first idea to series production. Thus, Michelin uses Siemens Digital Industries Software’s Plant Simulation in the Tecnomatix® portfolio to create a digital twin of its manufacturing facilities to virtually test changes and optimize processes before they are implemented, or to be able to react faster to the demand for changes or hazards in production.

This unlocks enormous savings potential for the entire company. Naturally, saving on resources impacts sustainability, another crucial part of Michelin’s future agenda, which is reflected in the slogan, “We manufacture the future,” which was introduced as part of the “All-Sustainable” strategy for 2030. This strategy, known as “Michelin in Motion,” aims to ensure the company can grow sustainably by balancing ecological, social and economic aspects.

Adopting the digital twin

The Plant Simulation digital twin has a far-reaching impact at Michelin, both internally and externally. Despite its importance, it could have remained a black box for many employees, accessible to only a few.

Thierry Chenevier and Aude Nursimiloo are familiar with this issue. Chenevier is one of Michelin’s digital transformation supply chain experts. Together with his colleague Nursimiloo, the project manager for digital twin manufacturing, they often encountered reservations and prejudices. Since end users had not been exposed to the digital twin, it was a challenge for the company to adopt it.

“That was our situation in 2022,” says Chenevier. “We had an excellent simulation model of our manufacturing, but unfortunately no broad user access, which we wanted to change. Michelin employs over 130,000 people worldwide, with about 80,000 in production. We have 86 production sites, about half of which produce end products. The simulation models we use are specifically for these manufacturing sites that produce end products.

“We had a vision for a one-click simulation experience that could answer questions like, ‘How does production capacity react if I change this or that parameter?’ We wanted people on site who were responsible for these issues to ask such questions. These are people from industrial engineering or planners and production managers. But the reality at that time was that we
had to adjust 65 tables for a new simulation, taking at least three weeks to prepare the data input. Moreover, the simulation was not accessible to everyone. Everything ran on the data analyst’s PC. Finally, there was no user-friendly interface for the nonspecialist user.”

Developing the digital twin

In 2018 Michelin began working with the digital twin in Plant Simulation for manufacturing. This phase generated a first simulation model for a site in Italy. Michelin was supported by Siemens partner Inoprod and tested the reliability of the technology and further possibilities. In 2021, they expanded this practice to four more production facilities, accelerating simulations and depicting application scenarios. The third phase has been ongoing since 2023, involving a more professional architecture of the digital twin and removing barriers to automation and providing a better end-user experience. Starting in 2024, five more sites per year are starting to use the Plant Simulation digital twin. This final phase focuses on training, automation, simulation model acceleration, simplification and massive expansion.

A game changer

“Plant Simulation offers a truly reliable and detailed foundation for innovation and progress,” says Chenevier. “This will provide added value for our manufacturing people who must have access. We are confident this will be achieved in our third development phase with the Mendix Optimize My Plant app. Optimize My Plant is a low-code software environment around Plant Simulation, which was a real game changer for us in terms of user experience.”

Plant Simulation and the Mendix™ Low-code platform are part of the Siemens Xcelerator business platform of software, hardware and services.

Optimize My Plant gains increasing acceptance

There are several examples of the how user-centric Optimize My Plant has gained increasing acceptance. In a curing shop, transferring the production plan into a schedule is a challenge due to many different presses and interlocking workflows. Manually creating it took between 4 hours and 2 days and in some cases an entire week. Today this is possible with one click in a comfortable overview. Complex data entry could thus be automated and accelerated.

Although in the past it could take weeks for key users in the plants to create new scenarios in Plant Simulation, Michelin has predefined scenarios with Optimize My Plant that any responsible person can run through in no time. The app shows the impact of changing a parameter in a scenario in key performance indicators (KPIs). If the end user wants to delve deeper into KPI analysis, they can retrieve details of the respective scenarios. This shows the development of production figures, inventory levels, possible stock shortages and much more.

It is still complex to build a dataset, but far less so thanks to optimizing the generation of complex data, automating data preparation and separating roles. The key users manage the complexity of regularly calibrating a full new dataset, while the end users compare scenarios and understand the flow behavior.

Success stories from various plants

Additionally, the increasing spread of Plant Simulation to other plants highlights how crucial the software is to the company’s success. These successes are also communicated by Michelin to further increase acceptance of the digital twin and advance the democratization of simulation across the company. Nurisimiloo cites several use cases:

“In China, the problem was there were two automated storage and retrieval systems in the north and south. One warehouse reached capacity when loading while the other still had plenty of space. Using Plant Simulation created a balance for our warehouses.

“In Italy, Plant Simulation was used at the start of a project to implement a new machine. The question was whether none, one or even two of the existing machines could be removed and how these scenarios would affect material flow. Using Plant Simulation made it possible.

“A Polish plant wondered how to better adhere to the production plan and avoid unnecessary press openings to save energy. They want to use Plant Simulation to enable them to achieve optimal press utilization, resulting in the desired energy savings.

“Maintenance managers in an Italian plant wanted to know how best to distribute work. Was it better to schedule it on a specific day of the week or spread the work over many small time slots to minimize impacts on production capacity? Using Plant Simulation delivered the reliable answer here as well.

“To increase production capacity, Plant Simulation is also being used to solve another type of time optimization for a plant in Poland: They are currently optimizing the distribution of the curing production plan on the presses based on the curing time.”

Expanding simulation technology

Given all these positive experiences, Chenevier is optimistic about achieving the company’s ambitious goals: “With Plant Simulation and Optimize My Plant as an easy-to-use frontend for our end users, we have realized our dream architecture. We tested this in five pilot projects, all of which strengthened our confidence. It is promising for future development. Moreover, we succeeded in convincing our staff of the value of dynamic manufacturing simulation for all areas of our industrial production.

“In the short term, we think current projects and everything related to optimization will benefit from this. In the long term, we hope to have a large-scale expansion of manufacturing simulation across all areas with real-time data retrieval. With Plant Simulation we are working on our vision of a seamless user experience. We are expanding our joint capabilities at all levels around this new productivity tool. Simultaneously, we are expanding our efforts for a full-scale implementation of Plant Simulation by this model.

“By 2027, we will have implemented this in 15 new plants. In the meantime, we are collecting more use cases and publishing them to increase interest and understanding of the benefits of making manufacturing simulation available to all relevant parties in our organization.”