Webinar Replay: Optimizing manufacturing supply chain with Digital Twins – a use case with Michelin

Overview

The webinar presents a Michelin use case for using simulation-based digital twins to improve manufacturing and supply-chain decisions. Unlike digital twins used only to monitor a physical asset, this approach models a broader system—including factories, inventory, logistics, demand, processes, and management rules—so teams can test possible futures before making decisions.

Michelin’s Supply Chain and Challenge

Michelin described a global supply chain with around 120 industrial activities, multiple manufacturing and distribution stages, and roughly 3.5 million square meters of warehouse space. The company cited a total lead time of about six months, annual logistics costs of approximately €1.9 billion, and nearly €4 billion in inventory.

The featured use case addresses a strategic question: Where should Michelin produce tires for the Chinese market, and which products should be supplied locally versus imported? The decision depends on factors such as factory capacity, demand and its variability, transport times, operating costs, product life cycles, and the cost of transferring production between plants. Michelin said the challenge has become harder as the number of differentiated products has grown.

Simulation Digital Twin Approach

Cosmo Tech’s simulation digital twin represents the supply chain’s structure and dynamics, allowing users to run “what-if” scenarios and compare potential outcomes. In the Michelin case, the model evaluated 80,000 sourcing strategies—a scale the speakers said would be impractical to assess manually.

The model’s outputs include:

• Profit and total cost
• Customer service levels
• Factory utilization
• Inventory levels
• Customs duties and transportation impacts, including CO₂ emissions

Michelin reported that customs duties were reduced by 60%, representing more than €10 million per year. The speakers also emphasized using simulations to assess uncertainty and risk before major investments, such as building a new factory.

Broader Applications and Integration

Michelin noted that digital twins can address different planning horizons and problems, including:

• Strategic and tactical sourcing across plants
• Flow optimization between plants, including analysis of the bullwhip effect
• Dynamic simulation of flows within an individual plant

The speakers said the model can connect with Microsoft Azure Digital Twins and IoT platforms. IoT data can provide information such as container locations, lead times, and stock levels; the simulation then uses the system model to explore future scenarios. They cautioned that IoT data is most useful when it is relevant to the decision being modeled.

The central argument is that a useful digital twin should not be limited to representing a single physical asset. For supply-chain decisions, it may need to represent the whole organization and the interactions among assets, people, processes, inventory, and operating rules.

Reviews, Guides, or Tutorials

No product review, step-by-step guide, or tutorial was presented. The webinar was a case-study discussion with audience Q&A.

Main Speakers and Sources

• Thibault (Tibo) de Rouville — Vice President, Group Industrial Supply Chain, Michelin; presented the supply-chain challenge and use case.
• Michel Morvan — Co-founder and Executive Chairman, Cosmo Tech; explained the simulation digital twin concept and its role in decision-making.
• Meg — Representing the Digital Twin Consortium; webinar host and moderator.