In pursuit of the efficiency of production lines, Trigo does not want to give up on automation. A group specializing in inspection and safety of production lines announced on Thursday the acquisition of startup Scortex, which specializes in quality control automation solutions for manufacturers.
This startup, launched in 2016, has made industrial innovation its battleground. His team has spent nearly six years building deep learning algorithms to automate validation and get real-time data and statistics on product quality. This time-consuming process requires large amounts of data so that the machine can sort out defective and conforming products, just as a human operator would.
Trigo, which primarily works with vehicle manufacturers in the automotive and aviation sectors, began experimenting with Scortex use cases a year and a half ago. Mathieu Rambeau, CEO of the Trigo Group, notes that the automation solution is particularly well suited to inspect “large and repetitive” parts such as bearing assemblies.
Reliability and repeatability
“Quality is not a black or white sphere, it is criss-crossed by large gray areas,” emphasizes Aymeric de Pontbriand, Founder and CEO of Scortex.
If the human operator is not infallible, then the controlled machine, on the other hand, is somewhat more flawless: “The Scortex solution allows subjective faults to be objectified with very high reliability and repeatability in detecting these faults,” explains Aymeric de Pontbriand. .
For Trigo, the automated tool has another benefit: the collection of information on a large scale that can improve product quality. Trigo offers industrial quality engineering and consulting services to its customers. Thus, such a tool in her purse could help the company “analyze the faults found and understand the causes, and thus offer solutions to solve problems for our customers,” explains Mathieu Rambaud.
Scortex, for its part, intends to double its workforce this year to speed up its product “go to market” and scale through the Trigo network. The startup also wants to continue its product roadmap and digitize operational quality management “to consider as many use cases as possible and provide state-of-the-art tools that allow experts to focus on real-time data,” says Aymeric de Pontbriand.
At a more advanced stage, AI also leads to prediction. According to Aymeric de Pontbriand, “We can start to detect weak signals if the quality drifts.” Thus, “before deciding that a part does not meet the requirements, we will be able to warn about it earlier. »