1. Reduce QC costs, unnecessary waste and increase throughput

Efficient, reliable vision inspection

Some of the most obvious costs that our automated inspection reduces are manual labour costs. Based on leading vision technology build on highly advanced algorithms, TriVision systems can reliably perform accurate quality control of all products in high speed productions and hereby release manual labour resources from tasks that increasingly challenge the human eye.

The accurate failure detection eliminates all falsely rejected products. With the ability to set up and accurately follow thresholds for all error types, TriVision systems will only detect and reject products with variations that actually are defined as defective products. As a result, unnecessary food loss, waste and costs will be avoided.

As our automated vision-based quality control can inspect products at high speed, multiple packages per second, operators can maximize the speed of their conveyor belt and overall production to increase throughput and hence OEE while still ensuring complete quality assurance.

A user-friendly system for operators to minimize costs

We have made sure that operators on the shop floor easily can utilize all the valuable data that TriVision systems generate when inspecting each product. The system brings a clear overview of production performance and error distribution with real-time data and sets off immediate alarms if an unacceptable number of defective products appear.

Preventive alarms can also be set up to spot and predict specific tendencies in the production, e.g. if label placement worsens. Based on inspection data and advanced mathematical models, the system can spot and predict if a tendency of e.g. slightly tilted label placements, but still acceptable placements, will worsen and result in unacceptable label placements. If an issue is predicted, the system can preventively alert the operator. This enables the operator to take preventive actions in terms of making adjustments or planning maintenance, before the tendency turns into unacceptable errors that cause food loss, waste, and unplanned stops.

Hence, the preventive alarms improves OEE significantly by reducing food loss, waste and converting unplanned stops into planned maintenance.

2. Avoid expensive recalls and sales decline

The purpose of quality assurance within the food industry is first and foremost to ensure the quality of products that leave the factory to reach supermarket shelves. Whereas quality assurance based on manual labour minimizes the risk of sending out defective food products and packaging, a TriVision system eliminates that risk with reliable inspection of all products and full product quality documentation. A risk that is worth eliminating, as errors in package seals, labels, date markings or appearance can cause bad customer experiences, sales decline, recalls, fines from retailers and in worst case food safety hazards.

Even if manufacturers would choose to reduce conveyor belt speed and decrease productivity to allow manual labour inspection of all products, humans are expected to make mistakes at some point. The number of defective products that will slip through annually and its consequences will vary for each manufacturer.

However, our experiences show us that a TriVision solution is a good business case for most manufacturers using manual labour inspection based solely on how our solution can eliminate the annual average costs of recalls and fines.

3. An efficient production optimization tool based on vision data

Additionally, manufacturers can utilize the generated inspection data further to perform production optimization. Data is generated from all inspections and is converted into elaborate production statistics in TriVision’s cloud-based Production Intelligence data base. As the production statistics are based on accurate vision data, Production Managers can get true, valuable insights into statistics on the overall OEE, uptime, rejected products, error type distribution and cavity error distribution. It allows them to potentially spot trends across various process parameters such as material suppliers or work shifts.

Based on the historic inspection data, our intelligent algorithms and mathematical models, TriVision’s Production Intelligence solutions automatically generate tangible recommendations based on these inspection data, which support managers in spotting the production trends that are profitable to change. Hence, with a TriVision system, site management can always count on having concrete suggestions for optimizing their production.

What is Machine Vision Inspection?

Machine Vision Inspection is a technology based on image processing that is used to automate inspection processes in production lines in various manufacturing industries. Basically, it allows a computer to inspect and evaluate product quality using industrial cameras, lights, and computer vision software.

The technology is used to perform high-speed, reliable quality control to ensure product quality and improve overall efficiency as it releases manual labour from repetitive and challenging inspection tasks and/or expands the existing quality control.

Additionally, by digitising the inspection process and generating documenting data, machine vision inspection can be used to provide production data and statistics, which can facilitate production optimisation.

Due to its wide application and efficiency, machine vision inspection is widely and increasingly used in several manufacturing industries such as automotive, electronics, food and beverage, pharmaceuticals, packaging, labelling, and printing.

Machine vision-based inspection is carried out by a machine vision system.

What does a Machine Vision System consist of?

A Machine Vision Inspection System consists of hardware components, the system’s body, and computer vision software, which can be considered as the system’s brain.

The Hardware Components

A Machine Vision system’s exact hardware setup depends on its specific inspection tasks and hence configuration, but basic components of a Machine Vision System count:

• Sensors that identify the presence of incoming objects
• Cameras, the eyes of the system, used for acquiring images of the
  objects that are to be inspected.
• Lights and often a housing to ensure optimal light for the image acquisition.
• A PLC for controlling the system logic i.e., inputs, outputs and other actions from hardware components.
• An industrial computer running the computer vision software that analyses the acquired images and provides a HMI.
• An operator panel that visualises data and enables operators to control the system.

The Computer Vision Software

The brain of the Machine Vision System, the computer vision software, should be perceived as a library containing several software packages that each provides a specific inspection type, i.e., enables the system to inspect and evaluate specific elements of an object.

It processes and evaluates the acquired image of the inspected object to answer the important question: Does the inspected object have any defects compared to the desired quality level set by the producer?

How does Machine Vision Inspection work?

When a machine vision system inspects an object, the following steps take place:

1. Image Acquisition

First, when the incoming object approaches the system, it is detected by a sensor. Light sources and cameras act on the sensor’s input by lighting up the area and capturing an image of the object at exactly the right time.

Nothing happens without a successful image acquisition. Using the right optics and lighting technique is key to ensure high image quality, which is fundamental for carrying out an accurate, reliable analysis and hence quality control.

2. Image Processing and Analysis

The acquired image is pre-processed by applying algorithms and a number of specific filters – to detect the product in the image and highlight the key features of interest. Based, on the filtered image, algorithms are then used to generate a data object – data containing the characteristics of inspected object’s highlighted features.

The design of an algorithm impacts the efficiency of the image processing and analysis. Powerful algorithms facilitate accurate, high-speed vision inspection, and hence maximises product quality and throughput.

3. Result evaluation

The system’s software compares the extracted data object, i.e. the characteristics of the object’s highlighted key features with a set of predefined tolerances, which defines a producer’s target quality level.

4. Decision Making

Finally, based on the comparison between the processing results and the set tolerances, decisions are made by the system. The inspected object will get a classification e.g. passed or failed. Some machine vision systems will be able to reject the product or initiate warnings, alarms, etc.

5. Process Control

Some machine vision systems additionally utilise the gathered inspection data to spot trends in the production, and give recommendations to the machine operators for production optimisation. This can be done via data platforms that provide production data overviews highlighting crucial production information.

An overview of production data can help manufacturers to understand why faulty products and errors occur and support them in identifying root causes. Also the adjustment of process parameters can be monitored and evaluated to optimise the overall production, and the performance of different shifts, production lines and production facilities can be compared to align and optimise production procedures.

Machine Vision Inspection in the Food Industry

Machine Vision inspection is an attractive technology for food producers to meet the high demands and cope with the challenges they face.

Not only are customers and retailers increasingly demanding high product quality and quality documentation – food producers simultaneously need to optimise to remain profitable, focus on sustainability and deal with labour shortages.

What can Machine Vision Systems check?

The application of machine vision inspection is vast, which means that any food producer in any sector can utilise the power of the technology to handle specific issues related to product quality.
Machine vision systems can be implemented into different production lines and configured to match specific application and inspection requirements.

Food producers can utilise machine vision inspection to perform:

• Inspection of food and packaging appearance
• Presence/absence check
• Orientation check
• Verification of amount and shape
• Colour inspection
• Foreign object detection
• Seal inspection and validation
• Label check and verification
• Reading and check of printed date markings
• Reading and check of barcodes and QR codes.

Machine Vision Systems are The Eyes of The Robots

As automation is key to increase efficiency and deal with labour shortages, producers are focusing on automating older existing production lines. Robots are replacing people in processing and packaging lines to increase throughput. And new production facilities are built fully automated today.

Hence, dark factories are the future, meaning that factories in theory would require no lights, as all processes are fully automated by machines.

It also means that new quality control solutions are required as producers do not have the same number of people on the production floor to monitor product quality, and increased production speeds due to an increasing automation level leave workers with inspection tasks that challenge the human eye.

This makes machine vision inspection a technology of the future. With its ability to inspect all produced products and detect the tiniest deviations and errors consistently and reliably over time at high speeds, enabling producers to meet the high demands of consumers and retailers while increasing efficiency.

More Machine Vision Insights

How to utilise the Machine Vision Technology

Even though machine vision holds great potential for boosting quality and production performance, it is a technology that is difficult to master and utilise consistently for efficient quality control. But if utilised properly, machine vision can boost a producer’s quality and production performance.

Read about how to achieve efficient quality control by using machine vision here

Discover all the benefits of Machine Vision Inspection

If utilised properly, food producers can implement machine vision systems in both food processing and packaging lines to improve product quality and documentation while reducing costs and facilitating optimisation of production processes.

Discover how Vision inspection in many ways can improve food producers’ business performance here.