UPDATED 11:00 EDT / JUNE 22 2021

AI

Google’s Visual Inspection AI tool automates quality control in manufacturing

Google LLC is catering to the growing demand for industrial automation with the launch of its new Visual Inspection AI tool.

As the name suggests, Visual Inspection AI (pictured) is an artificial intelligence service that relies on computer vision technology to automate quality control processes in manufacturing operations. It’s designed to be installed on machinery, where it can then scan thousands of products in order to identify any defects in them, before they’re shipped out to customers.

Implementing such technology sounds simple, but it really isn’t that easy. The problem is that manufacturers need to create customized vision inspection models for their specific operations.

So, for example, a computer chipmaker will need to create a purpose-built AI model that can spot defects within the processors it makes. That model must know what the products are supposed to look like and also what kinds of common defects might appear. Only then can it flag defective products as they work their way through the manufacturing chain.

Google said Visual Inspection AI simplifies that task. It requires no expertise to deploy, and that it can create capable models with as little as just 10 labeled images as opposed to the hundreds of thousands required by other AI services.

“Based on pilots run by several Google Cloud customers, Visual Inspection AI can build accurate models with up to 300 times fewer human-labelled images than general-purpose ML platforms,” the company said. “This allows the solution to be deployed quickly and easily in any manufacturing setting.”

Google also claims that Visual Inspection AI improves accuracy by up to 10 times compared with general-purpose machine learning approaches. In addition, the models it builds can detect multiple types of defects from a single image.

Another key benefit is that Visual Inspection AI is designed to create AI models with full edge-to-cloud capability, Google said. So this means the models can be downloaded and run autonomously on the factory machines themselves, helping improve latency and adhere to data governance rules.

Google said Visual Inspection AI has already been deployed by customers in multiple manufacturing industries, including semiconductors, consumer electronics and automotive. Defective circuit boards are a big problem in electronics manufacturing, with the average factory forced to throw away 6% of its products because of soldering errors, missing screws and the like, Google said. By employing vision AI to automate quality control, Google said, it could save a typical electronics manufacturing facility up to $23 million a year by reducing rework and material waste.

“We’ve been listening to the specific needs of the industry, and have brought the best of Google AI technologies to help address those needs,” said Mandeep Waraich, Google’s head of product for industrial AI. “The outcome is an AI solution that, built upon years of computer vision expertise, is purpose-built to solve quality control problems for nearly any type of discrete manufacturing process.”

Google said it’s working with partners including Accenture Plc, Siemens AG, SOTEC Software Entwicklungs GmbH and other to help customers in the manufacturing industry quickly integrate Visual Inspection AI into their operations. International Data Corp. analyst Kevin Prouty said Google’s approach to visual inspection provides just the kind of roadmap that manufacturing companies are looking for.

“Manufacturers want flexibility, scale, inherent edge-to-cloud capabilities, access to both real-time and historical data, and ease of use and maintainability,” he said. “Google is one of those companies that has the potential to bring together IT, operational technology and and ecosystem of partners that manufacturers need to deploy AI on the shop floor at scale.”

Visual Inspection AI is rivaled by Amazon Web Services Inc.’s Lookout for Vision service, which is similarly designed to be deployed on factory floors where it can spot defects in products. But it’s not just the public cloud giants that are competing here. A startup called CrowdAI Inc. recently burst onto the scene itself, raising a $6.25 million round of funding for its multipurpose platform that can be used to build vision AI tools for monitoring production lines and visually inspecting infrastructure.

Image: Google

A message from John Furrier, co-founder of SiliconANGLE:

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU