3 Reasons Why the In-Sight D900 Will Automate More Inline Inspections

In-Sight D900 deep learning vision system

There are many phases of a typical manufacturing process that go uninspected for one reason or another. Maybe there are steps during the inline process that can’t be automated because the inspection is too complex or challenging to program. Maybe the inspection requires some level of human cognition, expertise, or judgement.

Whatever the reason, when products go uninspected manufacturers open themselves up to product recalls, inefficiencies, rework, and other negative business outcomes that create headwinds towards increased profitability.

It’s estimated that a product recall can cost a company anywhere between $10 million and $100 million or more on the bottom line, but the lasting damage to a brand’s reputation and the slowdown in innovation could be far worse, says Harvard Business School professor Ariel D. Stern. As recently as 2016, recalls costs the entire automotive industry approximately $22 billion.

The cost for manufacturers and brands is too great to allow aspects of the production process to remain uninspected.

We’ve heard consistent feedback from our factory automation customers about how adding deep learning technology to the In-Sight platform could help solve these inline inspection challenges.

Fortunately, they no longer have to wait because the latest In-Sight vision system, the D900, was designed to help customers expand the types of inline inspections they can automate and scale.

The In-Sight D900 is the first vision system with embedded deep learning inspection capabilities allowing non-programmers to benefit from the transformative power of artificial intelligence (AI). Running In-Sight ViDi software, it solves your most challenging OCR, assembly verification, and defect detection applications without the need for a PC during deployment. The In-Sight D900’s combination of power, accessibility, and ease of use will help manufacturers inspect the previously uninspectable, while delivering fast, consistent, and accurate results.

There are three key reasons why the In-Sight D900 helps manufacturers and automation engineers across industries as diverse as automotive, packaging, electronics, or food and beverage.

1. Applications are easy to build and deploy

Deep learning jobs for the In-Sight D900 are built with In-Sight ViDi software, which takes advantage of the intuitive In-Sight spreadsheet interface, to quickly set up and run deep learning applications without programming. For In-Sight users, it’s the familiar programming environment you already know and love.

For those just getting introduced to In-Sight, the spreadsheet simplifies application development and streamlines factory integration with a full I/O and communications function set. It also allows users to combine traditional Cognex rule-based vision tools (like PatMax Redline) and deep learning tools in the same job, leading to quicker deployments.

In-Sight ViDi also requires vastly smaller image sets and shorter training and validation periods than other deep learning solutions. As a result, applications are quick and easy to set up, teach, and deploy to the In-Sight D900.

2.Deep learning is embedded on a powerful smart camera vision system

Until now, the use of deep learning for factory automation inspections required deep learning software, whether VisionPro ViDi or some open source framework, running on a PC and connected to a GigE camera, adding to the level of complexity to bring a deep learning solution to the line.

By contrast, once an application is created and the deep learning model is trained it gets deployed to the In-Sight D900 right on the line without a PC. That’s because the D900 has an embedded inference engine co-processor specifically designed to solve complex deep learning applications at production line speeds. And, like other In-Sight vision systems, it’s a highly modular, rated for IP67 applications, and includes field-changeable lighting, lenses, filters, and covers that can be customized to match your exact application requirements.

3. Automate and scale the most challenging inline inspections

The In-Sight D900 vision system with In-Sight ViDi software really shines when it comes to three specific applications: defect detection, challenging OCR, and assembly verification. With three new inspection tools, ViDi Detect, ViDi Read, and ViDi Check, this deep learning solution can be used for a wide range of applications. For example, on complex parts and surfaces where defects can be unpredictable in their appearance, ViDi Detect is a perfect tool for the job.

It also finds complex features and objects within an image or verifies parts and kits are assembled correctly based on their location within a user-defined layout. And when it comes to OCR, we all know machine vision works well with good lighting and clear contrast for the code you are trying to read. However, deep learning reads most codes under the most challenging environments. Whether a code is badly deformed, skewed, poorly etched, or on a reflective surface in terrible lighting, deep learning can handle it.


The In-Sight D900 is truly the first of its kind: a powerfully smart vision system capable of running deep learning inspections right on the line without the need for a PC. For too long, deep learning has been the provenance of academic theory and big tech companies. Now, though, automation engineers and quality managers can leverage its power right on the factory floor.

Get a Demo

More Posts on


Join MyCognex


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

Contact Us