{"id":2509,"date":"2022-08-18T16:39:05","date_gmt":"2022-08-18T16:39:05","guid":{"rendered":"https:\/\/salarydistribution.com\/machine-learning\/2022\/08\/18\/startups-vision-ai-software-trains-itself-in-one-hour-to-detect-manufacturing-defects-in-real-time\/"},"modified":"2022-08-18T16:39:05","modified_gmt":"2022-08-18T16:39:05","slug":"startups-vision-ai-software-trains-itself-in-one-hour-to-detect-manufacturing-defects-in-real-time","status":"publish","type":"post","link":"https:\/\/salarydistribution.com\/machine-learning\/2022\/08\/18\/startups-vision-ai-software-trains-itself-in-one-hour-to-detect-manufacturing-defects-in-real-time\/","title":{"rendered":"Startup\u2019s Vision AI Software Trains Itself \u2014 in One Hour \u2014 to Detect Manufacturing Defects in Real Time"},"content":{"rendered":"
\n

Cameras have been deployed in factories for over a decade \u2014 so why, Franz Tschimben wondered, hasn\u2019t automated visual inspection yet become the worldwide standard?<\/p>\n

This question motivated Tschimben and his colleagues to found Covision Quality<\/a>, an AI-based visual-inspection software startup that uses NVIDIA technology to transform end-of-line defect detection for the manufacturing industry.<\/p>\n

\u201cThe simple answer is that these systems are hard to scale,\u201d said Tschimben, the northern Italy-based company\u2019s CEO. \u201cMaterial defects, like burrs, holes or scratches, have varying geometric shapes and colors that make identifying them cumbersome. That meant quality-control specialists had to program inspection systems by hand to fine-tune their defect parameters.\u201d<\/p>\n

Covision\u2019s software allows users to train AI models for visual inspection without needing to code. It quadruples accuracy for defect detection and reduces false-negative rates by up to 90% compared with traditional rule-based methods, according to Tschimben.<\/p>\n

The software relies on unsupervised machine learning<\/a> that\u2019s trained on NVIDIA RTX A5000 GPUs<\/a>. This technique allows the AI in just one hour to teach itself, based on hundreds of example images, what qualifies as a defect for a specific customer. It removes the extensive labeling of thousands of images that\u2019s typically required for a supervised learning pipeline.<\/p>\n

The startup is a member of NVIDIA Metropolis<\/a> \u2014 a partner ecosystem centered on vision AI that includes a suite of GPU-accelerated software development kits, pretrained models and the TAO toolkit<\/a> to supercharge a range of automation applications. Covision is also part of NVIDIA Inception<\/a>, a free, global program that nurtures cutting-edge startups.<\/p>\n

In June, Covision was chosen from hundreds of emerging companies as the winner of a startup award<\/a> at Automate, a flagship conference on all things automation.<\/p>\n

Reducing Pseudo-Scrap Rates<\/b><\/h2>\n

In manufacturing, the pseudo-scrap rate \u2014 or the frequency at which products are falsely identified as defective \u2014 is a key indicator of a visual-inspection system\u2019s efficiency.<\/p>\n

Covision\u2019s software, which is hardware agnostic, reduces pseudo-scrap rates by up to 90%, according to Tschimben.<\/p>\n

As an item passes through a production line, a camera captures an image of it. Then, Covision\u2019s real-time AI model analyzes it. Finally, it sends the information to a simple user interface that displays image frames: green for good pieces and red for defective ones.<\/p>\n

\"\"<\/p>\n

For GKN Powder Metallurgy, a global producer of 13 million metal parts each day, the above steps can occur in as quick as 200 milliseconds per piece \u2014 enabled by Covision software and NVIDIA GPUs deployed at the production line.<\/p>\n

Two to six cameras usually inspect one production line at a factory, Tschimben said. And one NVIDIA A5000 GPU on premises can process the images from four production lines in real time.<\/p>\n

\u201cNVIDIA GPUs are robust and reliable,\u201d he added. \u201cThe TensorRT<\/a> SDK and CUDA<\/a> toolkit enable our developers to use the latest resources to build our platform, and the Metropolis program helps us with go-to-market strategy \u2014 NVIDIA is a one-stop solution for us.\u201d<\/p>\n

Plus, being an Inception member gives Covision access to free credits for NVIDIA Deep Learning Institute<\/a> courses, which Tschimben said are \u201cvery helpful hands-on resources\u201d for the company\u2019s engineers to stay up to date on the latest NVIDIA tech.<\/p>\n

Increasing Efficiency, Sustainability in Industrial Production<\/b><\/h2>\n

In addition to identifying defective pieces at production lines, Covision software offers a management panel that displays AI-based data analyses of improvements in a production site\u2019s quality of outputs over time \u2014 and more.<\/p>\n

\u201cIt can show, for example, which site out of a company\u2019s many across the world is producing the best metal pieces with the highest production-line uptime, or which production line within a factory needs attention at a given moment,\u201d Tschimben said.<\/p>\n

\"\"<\/p>\n

This feature can help managers make high-level decisions to optimize factory efficiency, globally.<\/p>\n

\u201cThere\u2019s also a sustainability factor,\u201d Tschimben said. \u201cCompanies want to reduce waste. Our software reduces production inefficiencies, increasing sustainability and making the work more streamlined.\u201d<\/p>\n

Reducing pseudo-scrap rates using Covision software means that companies can produce materials at higher efficiency and profitability levels, and ultimately waste less.<\/p>\n

Covision software is deployed at production sites across the U.S. and Europe for customers including Alupress Group and Aluflexpack, in addition to GKN Powder Metallurgy.<\/p>\n

Learn more about <\/i>NVIDIA Metropolis<\/i><\/a> and apply to join <\/i>NVIDIA Inception<\/i><\/a>.<\/i><\/p>\n

Attend <\/i>NVIDIA GTC<\/i><\/a>, running online Sept.19-22, to discover how vision AI and other groundbreaking technologies are shaping the world.<\/i><\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"

https:\/\/blogs.nvidia.com\/blog\/2022\/08\/18\/covision-visual-inspection-for-manufacturing\/<\/p>\n","protected":false},"author":0,"featured_media":2510,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[3],"tags":[],"_links":{"self":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts\/2509"}],"collection":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/comments?post=2509"}],"version-history":[{"count":0,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts\/2509\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/media\/2510"}],"wp:attachment":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/media?parent=2509"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/categories?post=2509"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/tags?post=2509"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}