AI in Product Development and R&D With Michelin

Use Cases & Projects, Tech Blog Lynn Heidmann

Applied mathematics and using data has always been a part of product R&D. At Michelin in particular, the R&D department has more than 30 years of simulation history, modeling everything from how molecules in materials interact with each other to how products interact, wear, and perform on the road, and more.  So the possibilities for applying AI in product development and in R&D are wide, but how can machine learning and product experts work together to make those applications a reality?

For the past five years, Michelin has been working on incorporating more machine learning into its processes for tire design and testing. Léo Dreyfus-Schmidt, VP of Research at Dataiku and head of the AI Lab, sat down with François Deheeger, Senior Fellow AI and Data Science at Michelin to talk details. 

 

"We've been working on trying to put some machine learning models behind [the data], and it's working ... Of course, there's a lot of work to gather the data, to make sure that we are using the right parameters, the right features, etc., but the concept is working. But then, there are a lot of questions regarding the quality of those models. If they can help us to, let's say, have new ideas, that's fine. But the target or the goal would be to homologate the product through those sort of models. It means that we go to ... any car manufacturer and we say, ok, our model is predicting this, and you can trust us. So that's the ultimate goal."

— François Deheeger, Senior Fellow AI and Data Science at Michelin

How to Infuse AI Into Product Development and R&D Processes 

In order to reach their goal — being able to homologate the product through machine learning models — Michelin is working on developing in four areas:

  1. Building better models: The team is not in a big data environment, and they have a lot of physics knowledge from experienced engineers, so how, given this constraint and these assets, can they build models that use the deep knowledge and expertise to move toward hybrid simulation for accuracy and acceleration?
  2. Defining what is a good model. This includes enabling trustworthy AI solutions for engineers through prediction uncertainty and exploitation domain assessment.
  3. Coming up with solutions. Engineers don't want predictions for all possible combinations — they're looking for solutions, so the team needs to provide a way to quickly assess the best tradeoffs inside a defined parameter.
  4. Answering the question: "Is this true?" Is the predicted performance related to the way customers are seeing the performance? This means assessing solutions' robustness with respect to real use uncertainty  — including manufacturing and usage — to predict real-life performance (moving toward digital twins).

Want more? Watch the full video on how Michelin is incorporating machine learning and AI techniques (including deep learning) into their product development and R&D processes.

You May Also Like

Beyond Text: Taking Advantage of Rich Information Sources With Multimodal RAG

Read More

Level Up: LLMOps and MLOps With Dataiku and Snowflake

Read More

Adoption of Generative AI in Retail & CPG: The Time Is Now

Read More

Scaling GenAI Initiatives: Insights From Aimpoint Digital

Read More