Skip to the content

OCR: applying a laser-like vision to manufacturing processes
personRich McEachran eventJan 29, 2020

OCR: applying a laser-like vision to manufacturing processes

Optical Character Recognition (OCR) is not a new technology, but advances in artificial intelligence will enable manufacturers to get more value out of it. 

Whether it’s white goods with faulty wiring or automotive parts with defects, product recalls can hit a company’s bottom line, its reputation and customer satisfaction. 

Smart manufacturing is helping to reduce the number of product recalls and eliminating the likelihood of human error affecting output quality. For example, Optical Character Recognition (OCR)-based vision systems are being used to automate inspection and traceability, preventing defects and identify instances of incorrect assembly and missing parts. But with a factory producing a number of lines at any one time, mistakes can still happen: parts can end up on the wrong line and being inadvertently used for the wrong white good or vehicle.

Even though the majority of these mistakes are caught by vision systems before the products have been shipped out of the factory, the fact that a wrong part ends up on the wrong production line in the first place can force a manufacturer to suspend operations and even scrap a batch of products altogether. Not only does this mean the manufacturer has to shut down and restart its factory but there’s also lost productivity and output to consider, as well as the cost of procuring replacement parts; in turn, this can lead to delays in shipping products to customers.  

OCR has been identified by manufacturing leaders and industry experts as one of the technologies that will benefit from the emergence of Industry 4.0. When combined with other technologies such as artificial intelligence (AI) and deep learning (DL), OCR has the potential to transform inspection and traceability. 

More than just lightening the paperwork load

In its rudimentary form, OCR can be thought of as machines in the 1800s that turned printed text into a machine-readable code, which helped blind people to read. It wasn’t until midway through the twentieth century – a time when computers were gaining momentum – that the ability to use OCR for data entry started to be realised. By the 1990s, OCR was in wide use and was helping businesses to manage paperwork and to capture and store old documents. 

One of the most common applications of OCR in everyday business is using Adobe Acrobat to turn text and handwriting into searchable and editable PDFs. 

In the manufacturing industry, OCR can be more than just a tool to improve administrative tasks or to digitise a company’s paper-based contract management operations. Given that manufacturers will often procure parts from different suppliers, keeping track of them throughout the production process is one of the biggest supply chain headaches. OCR can enable tracking to be more efficient.

Source: Adobe Stock

Nearly all parts that move through a factory are marked with serial numbers or 2D barcodes, typically using laser marking, etching or engraving. However, this isn’t without its problems. For instance, depending on the material being marked, etched or engraved, a serial number or barcode can wear off and the quality of the text can degrade, making it illegible.

Being able to read these serial numbers and barcodes is essential. What’s more, being able to capture and then store and process the information can have a positive impact on supply chain flow. 

OCR-based vision systems enable manufacturers to capture this information quickly, from the moment parts enter a factory to when products are ready to be shipped out. With the right manufacturing set-up and with vision systems in place at all stations, if a part arrives at a station it shouldn’t have, it can be automatically rejected.

Although such a set-up can capture this data at speed, with up to hundreds of parts being inspected every minute, the full impact of OCR-based vision systems can’t be truly felt unless there’s some deep learning involved. 

By using AI to train neural networks to recognise text and characters, vision systems can carry out inspection and traceability tasks with even greater accuracy. The more information it processes, the better the DL neural network become at identifying even the minutest of illegibility. 

Furthermore, when connected with enterprise resource planning software, such a set-up can be used to create alerts and flag any anomalies. Those at board level could even use this knowledge to inform future decisions regarding procurement and to improve operational efficiencies.

Source: Adobe Stock

A cog in the future of cognitive manufacturing

The OCR market will be worth $13.38 billion by 2025. A report from Grand View Research predicts that it will grow by a compound annual growth rate of 13.7% over the next five years. 

Looking to the future, AI-driven OCR-based systems will likely play a key part in cognitive manufacturing – where cognitive computing, the Industrial Internet of Things and advanced analytics are combining to revolutionise manufacturing processes like never before. 

According to Keith Mills, editor of Metrology News: “Cognitive manufacturing enables companies to put a laser-like focus on quality throughout the life cycle of a product’s development – from design through manufacturing and even after distribution when companies must ensure product quality through warranty and support programmes.”

“[This] approach improves yield, reduces overall warranty costs, and helps ensure customer satisfaction for the lifetime of a product,” he adds.

To enable comments sign up for a Disqus account and enter your Disqus shortname in the Articulate node settings.

Recommended articles

Real-time insight: connected factories and the future of manufacturing

On the road to a greener system of vehicle manufacturing

How lifecycle costing tools can boost customer relationship management strategies

5 key trends in Industrial IoT for 2020

5 reasons innovation is needed more than ever

Latest articles

Why automation could lead to mass unemployment and how a Universal Basic Income might be the only solution

CERN’s asset management supercollider

5 Tips on How to Innovate

Process Model Innovation: A vital part of Modern Dual Innovation

Change activism for the really agile enterprise


Science and technology are the principal drivers of human progress. The creation of technology is hindered by many problems including cost, access to expertise, counter productive attitudes to risk, and lack of iterative multi-disciplinary collaboration. We believe that the failure of technology to properly empower organisations is due to a misunderstanding of the nature of the software creation process, and a mismatch between that process and the organisational structures that often surround it.