Michelin: Tire & mobility giant leverages API-driven data automation and improves production resiliency by 98%

Founded in 1889, Michelin is a 132-year-old French company that has become one of the largest tire manufacturers in the world. The business manufactures nearly 190 million tires a year for nearly every type of vehicle, including bicycles, airplanes, farm equipment, heavy-duty trucks, motorcycles, as well as NASA’s space shuttle.

The tire giant has invested nearly €650 million annually to support its sustainable mobility innovations. And when it comes to using technology to drive that innovation, Michelin has put data at the center of its IT strategy.

“Data plays a role everywhere [at Michelin],” says Michelin’s Chief Digital & Information Officer Yves Caseau. “Our treasure is our IP about rubber and metal. In the past, we extracted knowledge from a lot of tests, which is about knowledge engineering and data engineering. Our mission now is to do what we call data-driven engineering and data-driven R&D.” Caseau, who is responsible for developing the company’s overarching data strategies, complying with all data security and privacy regulations as well as leading all IT initiatives to efficiently and effectively deliver tangible value to the business. The IT staff supports all lines of business from manufacturing and supply chain to HR, procurement, and more.

Challenges

For a global company like Michelin, running a supply chain of its scale and complexity is arduous. Caseau’s team launched a big initiative aiming to achieve 98% uptime for its business critical applications that would optimize the internal flow of materials through the supply chain and ensure consistent performance.

The TRW application, for example, is one of its key applications that calculates and optimizes the loading of tires in trucks. TRW was designed to mitigate possible shortages or overages of tires when making deliveries to customers and support the company’s overall sustainability goals to reduce its environmental footprint. Batch programs ran every morning at 6 a.m. to trigger a purchase order in the supply chain. If the batch did not succeed, every processing error had a huge impact on the transportation costs, resulting in potential lost revenue as well.

Michelin wanted a better way to bring data agility into SRE workflows and accelerate application recovery, without committing to a lengthy and costly replatforming or redesign project.

Solutions

The team deployed Delphix’s API-driven data platform that met the following requirements:

• Provides an immutable application data record in seconds, or at the transaction level, granularities for analysis and recovery

• Provisions data into root cause analysis environments to help teams recreate, troubleshoot, and fix application errors

• Limits the need for additional infrastructure

• Empowers application teams to access, refresh, branch, and share data using DevOps APIs

Results

Faster data availability for root cause analysis (RCA). Delphix captures a continuous, granular history of production data changes and delivers point-in-time data for the moments just prior to, during, or after an event—even for complex, integrated apps. As a result, Michelin’s SRE teams can access environments populated with data in the state that it was right before the system went down or data was lost or corrupted.

Delphix also allows teams to automatically restore or provision data to multiple parallel forensic environments across a range of time slices, which replaces slow, manual processes for data delivery that delay troubleshooting. Being able to recreate the right application state—including the data—is key to reproducing the problem, which is often the hardest part in the full incident resolution process.

Reduce application downtime from 2 days to 2 hours. Application downtime can be extremely costly, resulting in lost revenue and stalled production lines, and cause harmful reputational damage. With Delphix, Michelin can maintain its production system and ensure high availability of its TRW application by driving down mean time to repair (MTTR) and low failures (MTTBF).