The Top 5 Problems Vision Inspection Can Solve

September 2019

Challenging Your Vision Inspection

Before we look at the top five problems that vision inspection can solve, it is worth pointing out that although vision systems can be great for ensuring a quality-driven process, it is also very important to consider how these quality control checks will be challenged.

To make this possible, test pieces can be manufactured to mimic known defect criteria and to allow inspection tasks to be challenged on the vision application. Failure to include this in the quality control process could result in vision systems drifting out, which could be a result of numerous causes, such as a change in the light source or dust on camera lenses.

The Top 5 Inspection Areas

• Detection of Imperfections
• Measurement
• Validation and Confirmation of Operations
• Positional Accuracy and Robotics with Vision
• Product Tracking and Verification

Detection of Imperfections

Detection of imperfections within an automated process is probably one of the most widely used reasons for using a vision inspection system to ensure high-quality production.

In the plastics industry, using machine vision inspection systems for inspection tasks can vary from short mouldings to burn marks. The same image processing approach can be used in almost any industry where quality standards are set against a quality standard or benchmark. The vision system will use a field which determines the area it will be looking at. The pixel resolution will then determine the size of the defect or imperfection that can be detected.

Surface automated inspections can challenge even the most experienced vision experts and smart cameras. Quantifying what a defect is is paramount; a specification simply stating that a defect is “not allowed” is not enough. The size of the defect will determine the hardware that has to be used and will determine the resolution of the camera to be used.

This can be worked out using the simple calculation

Pixel resolution = Size of the field of view in the Y Direction (mm) ÷ Vision pixel count in the Y direction.

Measurement

The measurement machine vision system can be used to inspect for crucial dimensions on a part or assembly. This can vary from looking at pitches or the distance between holes to measurements of fluids in bottling lines. The same principle for the selection of the camera is shown above.

Validation and Confirmation of Operations

Many manual processes utilise work instructions to ensure processes are carried out systematically. Unfortunately, on occasion, part of the process may fail, and a manual process may result in an inspection or verification not being carried out. The vision inspection system can, therefore, be fixed above a workstation to inspect processes for whether a part has been removed from a tote bin or a screw has been loaded into an engine assembly. The complexity of these systems can vary, and it may be that a full vision system isn’t required, but a vision sensor would suffice. Companies such as Keyence can supply a wide range of solutions to cover most inspection applications, using powerful image processing to carry out the vision applications.

Positional Accuracy and Robotic Applications

The introduction of vision systems to robotics has enabled robots to be used not only for traditional applications in a simple repeat manner but is also now enabling robots to have a “set of eyes”. A six-axis robot from ABB with a vision system can now determine where it will pick up from now by looking for the part. A 3D vision system can be used to pick random parts from bins by telling the robot where the part is. Algorithms work out the angle and orientation of the parts to allow them to be loaded into machines in a much safer and repeatable manner. The robot can also present parts to a vision system for inspection or to allow for realignment and placement. Alternatively, the camera can be mounted to the robot itself.

The addition of vision systems to robotics gives additional flexibility by allowing the robot to move away from fixed position coordinates and adapt to variations in part size, location, and even colour.

Product Tracking and Verification

Product serialisation and traceability have become an essential requirement in many industries none more so than the medical device and pharmaceutical industry. The use of scanning devices for product tracking using 2D and 3D barcodes, along with other technologies such as embedded Radio Frequency Identification (RFID), ensures that the whole production cycle and product are monitored from the start and throughout the whole supply chain. Even in our everyday lives, companies such as Amazon, which could be viewed as being a warehouse and logistics company, are now seen as a technology company and have excelled in the market of product tracking.

The accelerated advancements in vision technology and its use are increasing. Part of this is down to other technologies being developed in parallel. Once a very specialised area, many vision system manufacturers have brought clever, simplified systems to the market, making it more affordable and user-friendly, making it a market that will continue to grow and develop. This will also ensure improved quality control of products.

Follow us on LinkedIn, Twitter & Facebook to stay updated

Read more about our automation here…

• Benefits of Automation
• The Benefits of Machine Vision Technology
• Robots & Cobots – What is the Difference?
• Robotic Automation
• What is a URS? Why is it Important?
• Cost of Industrial Automation
• Advantages & Disadvantages of Robotic Automation

Contact Us