Google Visual Inspection AI spots defects in manufactured goods
Google today announced the launch of Visual Inspection AI, a new Google Cloud Platform (GCP) solution designed to help manufacturers, consumer packaged goods companies, and other businesses reduce defects during the manufacturing and inspection process. Google says it’s the first dedicated GCP service for manufacturers, representing a doubling down on the vertical.
It’s estimated that defects cost manufacturers billions of dollars every year — in fact, quality-related costs can consume 15% to 20% of sales revenue. Twenty-three percent of all unplanned downtime in manufacturing is the result of human error compared with rates as low as 9 percent in other sectors, according to a Vanson Bourne study. The $327.6 million Mars Climate Orbiter spacecraft was destroyed because of a failure to properly convert between units of measurement, and one pharma company reported a misunderstanding that resulted in an alert ticket being overridden, which cost four days on the production line at £200,000 ($253,946) per day.
Powered by GCP’s computer vision technology, Visual Inspection AI aims to automate quality assurance workflows, enabling companies to identify and correct defects before products are shipped. By identifying defects early in the manufacturing process, Visual Inspection AI can improve production throughput, increase yields, reduce rework, and slash return and repair costs, Google boldly claims.
As Dominik Wee, GCP’s managing director of manufacturing and industrial, explains, Visual Inspection AI specifically addresses two high-level use cases in manufacturing: cosmetic defection detection and assembly inspection. Once the service is fine-tuned on images of a business’ products, it can spot potential issues in real time, optionally operating on an on-premises server while leveraging the power of the cloud for additional processing.
Visual Inspection AI competes with Amazon’s Lookout for Vision, a cloud service that analyzes images using computer vision to spot product or process defects and anomalies in manufactured goods. Announced in preview at the company’s virtual re:Invent conference in December 2020 and launched in general availability in February, Amazon claims that Lookout for Vision’s computer vision algorithms can learn to detect manufacturing and production defects including cracks, dents, incorrect colors, and irregular shapes from as few as 30 baseline images.
But while Lookout for Vision counts GE Healthcare, Basler, and Sweden-based Dafgards among its users, Google says that Renault, Foxconn, and Kyocera have chosen Visual Inspection AI to augment their quality assurance testing. Wee says that with the Visual Inspection AI, Renault is automatically identifying defects in paint finish in real time.
Moreover, Google claims that Visual Inspection AI can build models with up to 300 times fewer human-labeled images than general-purpose machine learning platforms — as few as 10. Accuracy automatically increases over time as the service is exposed to new products.
“The benefit of a dedicated solution [like Visual Inspection AI] is that it basically gives you ease of deployment and the peace of mind of being able to run it on the shop floor. It doesn’t have to run the cloud,” Wee said. “At the same time, it gives you the power of Google’s AI and analytics. What we’re basically trying to do is get the capability of AI at scale into the hands of manufacturers.”
Trend toward automation
Manufacturing is undergoing a resurgence as business owners look to modernize their factories and speed up operations. According to ABI Research, more than 4 million commercial robots will be installed in over 50,000 warehouses around the world by 2025, up from under 4,000 warehouses as of 2018. Oxford Economics anticipates 12.5 million manufacturing jobs will be automated in China, while McKinsey projects machines will take upwards of 30% of these jobs in the U.S.
Indeed, 76% of respondents to a GCP and The Harris Poll survey said that they’ve turned to “disruptive technologies” like AI, data analytics, and the cloud to help navigate the pandemic. Manufacturers told surveyors that they’ve tapped AI to optimize their supply chains including in the management, risk management, and inventory management domains. Even among firms that currently don’t use AI in their day-to-day operations, about a third believe it would make employees more efficient and be helpful for employees overall, according to GCP.
“We’re seeing a lot of more demand, and I think it’s because we’re getting to a point where AI is becoming really widespread,” Wee said. “Our fundamental strategy is to make Google’s horizontal AI capabilities and integrate them into the capabilities of the existing technology providers.”
According to a 2020 PricewaterhouseCoopers survey, companies in manufacturing expect efficiency gains over the next five years attributable to digital transformations. McKinsey’s research with the World Economic Forum puts the value creation potential of manufacturers implementing “Industry 4.0” — the automation of traditional industrial practices — at $3.7 trillion in 2025.