Michelin is a global organization well known for its high-performance tires, popular travel guides, and numerous innovations around materials, services, and connected solutions.
Many consumers recognize the French brand by the famous “Michelin Man” mascot, whose body is composed of stacked white tires. Back in the late 1800s, the Michelin founders had the foresight when they selected this futuristic humanoid, given the company’s present-day focus on using data to accelerate artificial intelligence.
“Getting more value from data is about artificial intelligence and machine learning,” Michelin Group CIO, Yves Caseau said. “AI is going to intensify the digital transformation because it increases the value that companies can extract from data.”
Michelin relies on AI to improve the efficiency of supply chains and facilitate customer interactions with its products. AI is also expected to augment the intelligence of human factory operators by managing complex manufacturing processes, which will enable employees to spend more time collaborating around activities that require human interaction.
“AI is going to open up boundaries on human activity because it’s a sponge for complexity,” Caseau said. “If AI helps you hide that complexity, then you can look at new frontiers. I view AI as yielding a new way of working.”
How Michelin is building its data-driven strategy
As a global entity, Michelin operates as a hybrid organization by providing an enterprise architecture that supports localization. While this technological setup spurs innovation regionally and within specific business lines, it creates data silos and challenges with data sharing.
“We have always gathered lots of data at Michelin, and decisions based on data are part of our culture,” Caseau said. “The data-driven challenge is to make data circulate everywhere, to break the silos, and to make it much faster.”
That’s why Caseau is focused on building the right foundation for a data-driven Michelin, including system-to-system integrations that enable data and services to be securely shared across internal teams and with external partners.
In addition, Michelin wants to automate the backend of AI, which requires large, continuous quantities of data and instant and near-infinite amounts of computing power in order to train models. By solving the management, flow, and sharing of data, Michelin can reduce errors and ensure data is ready for developing AI and training models.
Snowflake delivers a powerful data platform
To promote its data-driven strategies and accelerate business processes, Michelin requires a modern cloud data platform to provide a single, accessible, and secure data environment.
“The interesting part about Snowflake is that it’s a good example of what I call ‘AI for ops,’” Caseau said. “You get automated, scalable, distributed sharing of data with a service which is as low maintenance as possible.”
For its cloud data platform, Michelin needs three critical data components:
1. Centralized source for data: Due to its hybrid model, Michelin operates with separate data lakes and data intelligence platforms in each of its regional zones. This multiplicity of scale creates synchronization and sharing challenges that Snowflake solves with its cloud-agnostic layer.
Snowflake provides a single data experience where business units maintain local data management but break down their data silos. Regardless of cloud infrastructure, provider, or location, everyone has the exact same data experience. For example, Michelin’s global sales teams can share the same sales figures but use BI tools that reflect local sales practices, which are often based on country or regional practices.
2. Data sharing and distribution: Rather than share data via EDI and APIs, Michelin wants to enable secure data sharing, which removes the burden of copying and transmitting data. With Snowflake, live data is always shared from its original location. Data doesn’t move. Instead, a single version of the data exists, and data consumers are granted secure, governed access.
Michelin can also leverage Snowflake’s automation to share global data everywhere, coupled with specificity for different regions. This instant, frictionless, and secure sharing of live data can take place not only within Michelin but also between organizations, which means data can be shared externally with organizations such as supply chain partners and customers.
3. Data loading and management: Michelin is focused on transforming into a company that can react to events on a customer’s timeline instead of Michelin’s. To that end, its strategy is to move away from managing data processes in batch mode during off-peak hours and evolve towards an event-driven architecture with continuous processing.
With Snowflake, Michelin can gain complete management of data flows and a CDC (change data capture) streaming feature that enables faster performance and saves time by only loading data changes.
The future of AI and data at Michelin
Michelin views data and analytics as an extremely important part of its organizational journey. Today, Michelin uses data to better run R&D and manufacturing. Tomorrow, the opportunities are endless, especially as AI is further honed and developed using global data.
As Michelin continues to build out its data strategy, Snowflake demonstrates a path forward for delivering continuous performance and accessibility. By providing a single source of data and secure data sharing, Snowflake will empower Michelin to unify its global data and enable stronger collaboration—internally, with partners and customers, and perhaps one day with real humanoids. Learn more about Snowflake and Michelin.