Machine Vision Systems

Quality ensured from every angle

How could vision revolutionise your workplace?

Measure critical dimensions on the fly

Track and rectify misalignment

Detect incomplete and damaged products

Target rejects with instant extraction

Predict maintenance and optimise production

For a fast, flexible and consistent inspection process, you need a tailormade vision system crafted by one our experienced engineers.

Measurement

vision inspection sick sensor

Contactless inspection of critical dimensions can be achieved often without any breaks in production. The measurements can be used to uphold quality standards, support preventative maintenance, provide closed loop control, or guide later steps in the assembly process.

Almost every industry can benefit from inspection measurement systems with the technology able to take recordings such as 3D distances, concentricity, radii and angles.

Examples of where such a system can be used:

Food & Beverage Production
Automotive
Pharmaceutical
Industrial

 

Presence Detection

The completeness (missing components, correct position, orientation) and quantity of items can be verified at before shipping or at individual stages of assembly. This means incomplete products can be apprehended before further processing and more importantly reaching the customer. Records can also be kept verifying the product status to identify root causes of product complaints.

Example uses:

Automotive

Food packaging 

 

vision jar inspection system

Location Tracking

vision data logging

This type of vision technology is commonly integrated with pick and place systems however it can also be used for sorting products and aligning for processing.

Incorrect positioning can be detected, allowing process equipment to adjust accordingly to rectify or remove the product. This can reduce machine jamming, protecting key equipment wear and reducing downtime.

Location tracking also allows for the prediction of future motion changes. We can program the process equipment to react appropriately, preventing problems before they occur.

For more complicated tracking tasks, 3D point clouds can be captured and matched to CAD models to identify the version, coordinates and orientation of a part. This data can be used to identify pick points for robots, evaluate complexed geometries and measure distances without a known camera position.

 

Defect Detection

When products do not meet the standards set, the quality of the product decreases.

Defect detection is an ideal solution for those looking to detect anomalies in their product and remove them from the process line.

Defects such as size, colour variation, unwanted marks or scratches can be detected. Furthermore, once defects have been identified, they can be analysed. This gives you a greater insight into the quantity and severity of defects that occur and where they occur.

 

vision data logging

Reject Systems

Reject system for production line

A vision system that can detect when an item is out of specification can be a vital component to your inspection process. However, once you’ve identified a substandard product, how can it be isolated and removed from the production line?

Reject Gate

Depending on your preference, this method can be intrusive (using a divert gate) or non-intrusive (using a pneumatic blower). The positioning of a reject gate can vary. You have the option for it to be either on the side or the bottom of the process line.

Push up Reject

This is where the defective item is pushed upwards onto a perpendicular reject conveyor separate from the main process line.

Pick up & Remove 

Using a mechanical device (such as a robot, linear actuator) the defective item can be removed completely from the process line.

 

Deep Learning and AI

Machine Learning has an important part to play in an effective vision system.

There are various factors that can impair a vision system’s performance. One key consideration is the lighting of the subject. This includes the ambient lighting variance, product reflectance and machine positioning. In many cases designing a custom light studio can overcome issues caused; this however is not always the most practical or effective solution.

When situations like this occur, we use deep learning (AI) to train neural networks which can identify patterns that ordinary vision cannot quantify. The technology is tuned to analyse ‘good’ and ‘bad’ images from your process. Our system will then use these classifiers as a guide when performing quality/inspection checks. The AI is trained robustly so that in changing conditions, the decisions made are the right ones.

Vision capture of inspection system.

Adbro controls vision project procedure

Inspection system for food production line

 

We pride ourselves in being able to offer you bespoke vision systems that will meet your exact requirements. With so many years in the process industry, we have the knowledge and experience to consult with you as to what is best for your business. 

After contacting us with your requirement, a dedicated vision engineer will get in touch to discuss  your needs in further detail. To enable us to advise the best solutions for you, a site visit can be scheduled where we can observe your process in situ. 

In order to assist with the development side of the project, where possible we like to obtain product samples to help aid tests carried out in house. These tests will  include the use of a range of equipment including cameras, lenses and lighting.  

Results obtained by these test studies will enable us to determine the best solution for your needs and allow us to advise the associated costs.

More details can be found on our Design and Build page