SenSigma: Smart Optical Monitoring Systems

Defect formation is a highly problematic part of the manufacturing process. Every year, faulty products cause billions of dollars in material costs and production delays.

U-M Mechanical Engineering and Materials Science Professor Jyoti Mazumder first became interested in online detection of defects for car body fabrication more than 12 years ago during a research project for Aetna Industries and, subsequently, for Toyota. As he explains, “Most car companies rely on a spot lap welding technique that makes it difficult to perform online diagnostics of the weld quality. This is because the weld pool at the weld interface is between two layers, denying any direct access for inspection. The only way to check the quality of a weld is to stop the line, perform detailed tests and, if breaks are discovered, do a massive recall.” Many car companies are now moving to laser welding for improved joint strength and lighter cars, since laser welding needs much less space compared to spot lap welding and thus a reduced amount of steel.

A specialist in laser materials processing and optics, Mazumder and his team developed Smart Optical Monitoring Systems (SOMS), which analyze the wavelengths of light emitted by laser-induced plasmas in real-time, during
the welding process. Based on that information, SOMS can detect the defects, elements present and quality of each weld. When installed on a welding robot, the SOMS identifies and categorizes defects and also can be programmed to “teach” the host welding equipment to create stronger, more cost-effective welds and material depositions.

In 2010, Mazumder teamed up with U-M Tech Transfer staff to create a business model. SenSigma was launched in 2011, and by 2012 the company had set up shop in the U-M Venture Accelerator. Currently, Mazumder is negotiating with robotics companies to integrate SenSigma technology with welding robots. Given the size and unmet needs of the four-billion-dollar welding market, the future looks bright for the startup.

An average mid-sized passenger car contains as many as 4,800 spot welds, each one a potential problem. Faulty welds cost American automotive manufacturers as much as $400 million annually. SenSigma’s Smart Optical Monitoring System (SOMS) can detect weld and material deposition problems in real-time. Able to analyze any weld that generates light, SOMS will help manufacturers reduce cycle time and waste, decrease vehicle weight and improve fuel efficiency.

[source: U-M Tech Transfer 2013 Impact Report]