What Computer Vision App Will Have the Greatest Impact on Industry 4.0

Curious what the leading minds in AI and computer vision think will make the most significant difference in the future of industry 4.0? Follow along as alwaysAI’s staff share their favorite computer vision (CV) applications for Industry 4.0.

What CV application, use case, or model are you the most excited about regarding its implementation in the development of Industry 4.0? 

Dalton Varney, Corporate Strategy & Business Development Manager:

I’m the most excited about new computer vision technology that’s centered around improving the retail experience.

Computer vision solutions can alleviate supply chain issues found in construction, semiconductor manufacturing, and more, in the manufacturing industry. Specifically, AMRs with objective avoidance and object detection enable use cases that don’t require constant monitoring by a human. This autonomy can give employees more time, energy, and money to spend on tasks that require human thought and action.

In the retail industry, computer vision-based checkout systems can improve the customer experience by streamlining the checkout process and reducing the time spent scanning each item. Most retail stores are still using antiquated technology such as barcodes and individual item scanning. As such, consumers spend a lot of time scanning items one by one after picking them off the shelves. If they scan more than one item at once, or move too quickly, they are also likely to encounter errors such as “please place your item in the bagging area.” This frustrating user experience can be resolved by computer vision applications that follow shoppers in the store, monitoring what is placed in their cart. These object identification and people tracking models allow shoppers to leave the store without the need for a formal checkout process.

Steve Griset, Co-Founder & CTO:

I'm most excited about three major applications of computer vision in Industry 4.0: vision-guided robots on the warehouse floor, defect detection and reduction, and barcode scanning.

Vision-guided robots on the warehouse floor

These robots are equipped with fiducial markers such as AprilTags, which are reference objects placed in the camera’s field of view when a video or image frame is captured. The computer vision software will then identify the AprilTags in the image or video and perform actions based on the tag, such as:

• Camera calibration
• Object size measurement
• Autonomous navigation
• Object orientation
• Distance measurement

 Defect detection and reduction

Computer vision applications can be a cost-effective solution for reducing and detecting defects during the product inspection process. A computer vision application can scan the product from several angles to match it to the acceptance criteria. It can also save the accompanying metadata, which can be used to inform the manager or stop production if the number of faulty products reaches a certain threshold.

Computer vision applications can inspect the following aspects of an item:

• Whether the edges are intact
• Whether it’s filled to the necessary level (if it is a container)
• Whether it has the desired color, length, or width
• Barcode scanning

Finally, like Dalton, I’m excited about computer vision applications’ barcode scanning abilities. We can simplify the checkout process by using computer vision applications to scan barcodes quickly and effectively on a large scale for manufacturers. This can:‌

1. Improve the manufacturer’s inventory management process
2. Speed up order processing
3. Enhance a tracking system

Marty Beard, Co-Founder & CEO:

Like Dalton and Steve, I’m very excited about what the future offers. Robots and computer vision will augment human operators in the warehouse, factory floor, and more.

The trend towards automation will only accelerate. It’s been accelerating already, especially in recent years, due to the enormous labor and supply chain disruptions we have recently experienced. 

I foresee CV being used in the following scenarios:

• To help robotic arms “see” and scan for anomalies and defects in production. This will help cut down costs by avoiding costly restarts.
• As an “auditing” function to make sure environments are safe for workers.
• To produce records of what happened if something goes wrong.

It should be worth noting that CV won’t outright replace human operators. CV and robotics will be integrated into the workplace so human operators can work more efficiently and effectively. CV will also help companies reduce costs by avoiding costly restarts, ensuring compliance with health and safety practices, and more.

Scott Miller, Head of Product & Partner Management:  

What am I the most excited about? How Industry 4.0 will digitize the factory.

Essentially, if you don’t digitize your factory, you haven’t done Industry 4.0. To digitize your factory, you need to build an activity graph for your factory, which I’ve called in a past life “the industrial graph.” Your factory’s machine parts, your factory’s location, and other physical factors like where cars are located, are vital parts of the industrial graph since digitization requires knowing how things are moving through at different times of the day.

Once you’ve figured that all out, computer vision will help you visualize and understand how people interact with different things in the industrial plant. You’ll find the answers to essential questions like:

• Who’s near what?
• What are these people doing?
• Where are these people located, and where have they come from?

By integrating computer vision into your workplace, your industrial graph will expand beyond just the things in your factory to include people and activities.