Minitab: when data and AI reinvent industrial performance

More than fifty years after its creation, the company intends to go beyond its image of specialized software to more broadly support the digital transformation of manufacturers. Sami Chaabi, Solutions Architect at Minitab, details this evolution and how data can become a real performance tool. Manufacturers have been accumulating data for years.

Where now lies the real issue? The subject is no longer so much about having data, but about being able to use it correctly. For our customers, they are often very numerous and above all decompartmentalized: some come directly from the factory, others from research and development or even from the supply chain.

We must succeed in collecting and analyzing them in a coordinated manner to obtain a global vision. This is particularly important in an environment where manufacturers seek to produce better with their existing resources and to control their costs. Minitab historically works with quality and operational excellence departments, precisely because their role is to improve processes and industrial performance.

Minitab is historically known for statistical analysis. How has your job evolved? Statistical analysis remains at the heart of our history, but we have been undergoing a significant transformation for around seven years.

We have gradually enriched our approach because before analyzing data, we must be able to collect it, in particular directly from the machines present in the factories. The acquisitions made in recent years now allow us to go much further: collect data, analyze it, monitor production and performance in real time, but also simulate an industrial process. This simulation makes it possible, for example, to assess the consequences of a machine breakdown or a lack of personnel even before the situation occurs.

We are thus moving from an essentially analytical logic to a much more global approach. Concretely, how can this approach improve the performance of a factory? The advantage is to be able to understand more quickly what is happening and to anticipate more.

Simulation is particularly interesting: a company can virtually test a new process or an evolution of its production before actually implementing it. Take the example of a pharmaceutical manufacturer who must produce a new vaccine. Before deploying the entire industrial process, it can simulate its operation, identify potential difficulties and measure the consequences of different scenarios.

The challenge is therefore to gradually move from data which allows us to see what has happened to data which helps us decide what we are going to do. What place do you give to artificial intelligence in this development? We integrate it sparingly: Minitab is not intended to create artificial intelligence for its own sake.

Machine-translated from French