Artificial Intelligence Helps Boost Manufacturing Capabilities

people discuss deep learning in an industrial setting

For decades, automation using machine vision has been one of the most popular ways for manufacturers to increase their margins. Today, automation via Artificial Intelligence (AI) technology is transforming manufacturing’s ability to improve their business operations and gain new customers.

At its most basic, AI enables machine and computing systems to learn from data and examples in order to predict outcomes. According to Forrester Research, an astounding 53% of technology decision-makers are either implementing or expanding their use of AI, with another 20% planning to implement AI in the next 12 months. Manufacturing is one of the largest industries to embrace AI, alongside healthcare and retail, with spending on AI in manufacturing projected to grow by nearly 50 percent per year, to hit $17.2 billion by 2025.

AI can be leveraged for areas as diverse as supply chain management, quality testing and inspection, or predictive maintenance for equipment. AI has a truly transformative power to recast the way manufacturers think about their entire operations. But even as this game-changing technology advances and becomes more user-friendly, many manufacturers still struggle to take full advantage of it, largely due to perceived challenges involving cost, startup time, required expertise, and the reliability of results.

By redefining their expectations of performance, whether that is defects caught, false rejects avoided, or time saved, manufacturers who embrace AI, and specifically deep learning-based applications, as part of their inspection automation strategy can enjoy major economic and material gains.


A successful deep learning project can generate cost savings, as well as yield improvements and a better understanding of your own manufacturing process. While there are initial, direct costs associated with implementing a deep learning solution—including software and hardware expenses, development costs for engineering personnel, and the time required to collect data inputs—the direct and indirect benefits are substantial.

Here, we explore three of the expected as well as unexpected benefits of AI beyond the direct financial ROI calculation.

Cut Costs and Reduce Overhead

Manufacturers willing to take the risk of replacing outdated work practices—specifically manual inspection, where machine vision is too difficult to implement—will be rewarded with less overhead. Manual inspection is dominated by labor costs, incurred yearly, and include staff turnover and re-training expenses. Human inspectors are frequently superior to automated solutions when they are paying undivided attention. But most operators can only focus for 15-20 minutes, resulting in inconsistencies during a shift or between production lines. When computing payback on an AI project, many manufacturers are surprised to learn how quickly their yield and throughput improve.

Quicken Implementation

This may sound surprising, since many people assume that the startup time involved in scoping an AI application is significant.  But new, easy-to-use AI software designed for factory automation can actually quicken time to market. Consider the time and effort involved to accurately program and maintain complicated machine vision applications that feature an element of human judgement: the defect libraries, the exceptions to account for, and filters can become immense over time. Instead of writing algorithms or programming complicated rules for a computer, AI teaches the same system to learn from data sets and make decisions based on those examples. With the help of a few quality engineers and a few hundred to a few thousand training images, an AI application can be implemented, tested, and refined in a matter of weeks.

Improve Analytics and Upstream Process Control

An AI solution that documents inspection results provides reassurance for its user, as well as the ability to retroactively check inspection images and decisions in the event of future failures. Once a final inspection station has been successfully automated, it’s often possible to migrate the inspection steps upstream to in-line inspection. This cuts costs by identifying defects sooner, before expending time or adding additional value to bad parts. Finally, deep learning-based machine vision can be tied to your overall process improvement initiatives, such as correlating concrete vision data to other metrics such as process recipes, component suppliers, equipment differences, factory location, and more.

AI is the new labor-saving automation technology to help manufacturers achieve additional margins and enjoy indirect benefits across the supply chain. By treating AI as part of an organization’s overall strategic automation plan and operations, this technology can help manufacturing companies achieve new performance heights, increase shareholder value, and pull ahead of its competitors.

To learn more, download Cognex’s free whitepaper, “Deep Learning for Factory Automation.”

More Posts on


Join MyCognex


Cognex representatives are available worldwide to support your vision and industrial barcode reading needs.

Contact Us