While the debate over the value of artificial intelligence in the supply chain rages on, at least one company believes it’s the basis for “reimagining manufacturing from the ground up.”

Bernard Mariette says AI holds the key to development of “fully autonomous” technology for precision manufacturing — essential to industries such as aerospace, automotive and medical components.

Mariette is executive chairman of Watch Out, the French/Swiss creator of a manufacturing system that claims to cut costs and carbon footprint by at least 30%, while drastically reducing the number of human workers required on the assembly line.

Precision manufacturing involves the machining of parts with accuracy of two to three microns. (One micron is 1/70^th the thickness of a human hair.)

The state of the art in industrial manufacturing has been based on the use of computer numerical control (CNC) machining – employing pre-programmed software to dictate the operation of tools and machinery — since the early 1980s, Mariette notes. The technology has, of course, improved steadily over the years, but he asserts that AI-driven autonomous manufacturing has “leapfrogged” everything that came before.

Mariette calls his company’s system a “microfactory,” able to carry out multiple processes, from computer-assisted design to invoicing and quality checks. It all happens, he says, “without human intelligence.” A series of sensors controls all tools, parts, temperatures and other environmental conditions that are critical to sensitive precision manufacturing.

The system allows for the setup of, say, an aerospace workshop in just two weeks instead of the traditional three to six months, Mariette claims. With it, an operation that previously required 50 highly trained workers now runs with just 12 relatively non-skilled operators.

Watch Out’s technology has been in development for a decade, although it’s just now hitting the global market. It’s in use in the U.S. and Europe by LISI Group, a supplier of fasteners and assembly components to Boeing, Airbus and other aerospace, automotive and medical manufacturers.

Mariette believes autonomous precision manufacturing has major implications for companies looking to reshore production to the U.S., where high labor costs and low availability are a challenge. “We started this project because we couldn’t find qualified labor,” he says. “No kid wants to work anymore in the factory. It was getting more and more difficult to find people.”

The situation was exacerbated by the COVID-19 pandemic, which accelerated the retirement of an older generation of skilled workers. “We can see every day that we’re missing people,” Mariette says.

An automated operation carries the additional advantages of enabling round-the-clock operation and creating higher-quality parts, with sensors and AI-based algorithms correcting errors in real time, he says.

The adoption of such sophisticated technology presents some significant challenges. It entails a substantial up-front investment, along with the need to integrate cutting-edge technology into legacy systems. In addition, retraining is required for the remaining humans who will be responsible for overseeing and maintaining the machines.

“The skill isn’t on the [plant] floor,” Mariette says. “It at the center — to analyze and make sure they’ve improved.”

He claims that the initial cost of the system is offset by the savings derived from being able to run it constantly, with fewer people. “It’s all about ROI, at the end of the day.” In the end, he adds, “Compared to buying a traditional state-of-the-art CNC [system], it’s practically the same.”

Watch Out is currently raising funds to expand its presence in the U.S. and European marketplaces. Mariette says sales doubled in the last three years, and are expected to do so again in the next three. The company is aiming for revenues exceeding $70 million by 2026.

Still to be determined is the ideal way to sell the system. One is to market it as a traditional capital expenditure; another is to offer it through s subscription.

“It’s easy to do this as a service,” Mariette says. “We’re testing it at the moment.”