No two potato harvests are the same, yet each of the 100 million Pringles cans that leaves the Kellanova plant in Kutno in central Poland has to have the crunch and flavor that consumers expect. The process of ensuring that consistency is increasingly guided by artificial intelligence.

Sensors, lasers and cameras feed data on everything from humidity to protein content into software from Siemens, which continuously adjusts the recipe for changing raw materials before quality problems can slow production or lead to waste. It’s an example of the promise of industrial AI, a branch of the technology where Europe still has a chance to play a leading role.

While the region has fallen behind the U.S. and China in consumer AI, it has a deep trove of production and manufacturing data and expertise from an industrial sector stretching back more than a century. That gives Europe an opening to develop technology that could eventually automate entire factories to make stuff people need, something chatbots can’t. And a transition to data-driven industry could chart a path forward for its embattled manufacturers in the process.