Everything in Sight: Intel and Shanghai DeepSight* Power Smart Manufacturing Success

Manufacturing success comes from greater visibility and deeper understanding of every part of the process, no matter how small. For many companies, innovations like computer vision, data analytics, and robotics systems have become integral to factory operations, creating new efficiencies that are disrupting existing value chains. Small, focused improvements are driving business transformation at an exponentially larger scale, and the barriers to deployment of AI-enabled Industrial Internet of Things (IIoT) solutions are constantly being lowered.

Even so, technological innovation can often outpace humans’ ability to conceptualize – and thus, embrace – widespread change, particularly in historically conservative industries like manufacturing. As my Intel colleagues Irene Petrick and Faith McCreary point out in their study of 145 participants from 133 manufacturers, “to successfully navigate the transition [to Industry 4.0], you must be able to translate the grand vision into changes that workers care about, namely solving the problems they face today.”¹

In other words, how can the so-called “factory of the future” make work easier and more efficient right now? What actually happens when the rubber meets the road?

This is where real-world success stories can be both instructive and inspiring. At Intel, in collaboration with industry-leading partners, we’re making the promise of Industry 4.0 more attainable by delivering breakthrough solutions to real business challenges.

How Computer Vision is Transforming Defect Detection in China

When a leading tire manufacturer needed a more accurate, more efficient way of identifying product defects, they looked to Shanghai DeepSight* for a custom computer vision solution. Leveraging its AI expertise – the team originated at Intel labs – and experience with industrial collaborations, DeepSight* built a detection tool that integrates seamlessly with the manufacturer’s operations and scans for specific types of defects.

The process begins with the imaging system capturing visual data of the production line. The data is then transmitted to an Intel®-based industrial computer for real-time analysis by DeepSight’s* defect detection software, optimized using the Intel® Distribution of the OpenVINO™ toolkit. OpenVINO™ provides computer vision and deep learning inference tools, helping maximize performance of Intel hardware. The classifications are then sent back to the production line, ensuring that defects are spotted early, in near real time. From start to finish, this new and improved process takes less than one second.

Consider the transformative effect the solution has had on this company’s defect detection process. Like most manufacturers, it previously relied on manual inspection, which has proven to be time-consuming and unreliable. Inspector training typically takes at least three months, with up to 80% of trainees’ time devoted to inspection. Once fully trained, humans continue to make subjective judgments and rarely go on to exceed a 95% defect detection accuracy rate.

Now, using the Intel®-based DeepSight* solution, the manufacturer’s accuracy rate is higher than 99.9% – inspecting more than 20,000 tires per day! These small time savings have steadily accrued into big cost savings, helping reduce labor costs by approximately $49,000 per production line. Through early detection, they’re also reducing waste and environmental impact, while helping increase overall product quality, deployment speed, and profitability. When you’re producing millions of tires per year, every added efficiency, no matter how small, can have a big impact on the bottom line.

Inspectors who were straining their eyesight and battling fatigue trying to keep up can now focus on value-added tasks requiring uniquely human skills, thus improving employee productivity and satisfaction. Ever-evolving predictive decision-making enables the machines to monitor and fix themselves, reducing maintenance and costly downtime. And built-in scalability and expedited learning has allowed the manufacturer to start applying the solution in new defect detection cases. In this factory and for manufacturers everywhere, AI-enabled technologies are awakening new possibilities every day.

OpenVINO™ and Intel® Edge Insight Software have since been incorporated into additional Industrial IoT solutions, which promise to create new efficiencies and drive further transformation in smart manufacturing. AI Ready Vision Kits from IEI* and Advantech* are designed for fast deployment of machine vision workloads to enhance factory automation and improve defect detection on highly detailed products, such as textiles. Solutions like ADLINK’s* AI-Enabled Machine Vision Kit will empower integrators with faster, simpler ways to solve critical challenges, helping to keep production humming. More innovations are coming, and the best will propel more manufacturers to business transformations of their own.

Rather than attempting to sell its employees on sweeping changes and pie-in-the-sky technologies, this tire manufacturer isolated a real, immediate problem and implemented a solution with real business value. The difference is, in the connected factories of Industry 4.0, small benefits compound over time and reverberate throughout the workplace to bring big, transformative improvements. With the help of Intel and partners like DeepSight*, the grand vision of intelligent manufacturing is being realized, one fraction of a second at a time.

Visit marketplace.intel.com for more information on DeepSight* AI Labs.

___________________________

Disclaimers

1. Industry 4.0 Demands the Co-Evolution of Workers and Manufacturing Operations

Intel, the Intel logo, and OpenVINO are trademarks of Intel Corporation or its subsidiaries in the U.S. and/or other countries.

*Other names and brands may be claimed as the property of others.

Published on Categories Internet of ThingsTags , ,
Christine Boles

About Christine Boles

Vice President, Internet of Things Group - General Manager, Industrial Solutions Division - Christine's organization is responsible for Intel’s Industrial IOT business within the manufacturing, energy, logistics and commercial building segments, including the product and ecosystem strategies for this rapidly evolving space. Boles joined Intel in 1992 as an application engineer for 16 bit microcontrollers. For 25 years, she has led development, delivery and the enabling of customers and ecosystems for Intel based solutions in many managerial roles. These solutions span a broad range of embedded and internet of things applications across many industries, including communications, storage, retail, imaging and commercial buildings. Boles holds a Bachelor of Science in Electrical Engineering from University of Cincinnati and an MBA from Arizona State University.