Crush Defects: The Power of Semi-Automatic Inspection

Posted on

This month, we’re discussing the elements of semi-automatic inspection. This is the halfway house that pharmaceutical manufacturers have to move to from manual quality inspection (of individual products) at a bench when volumes start to increase but don’t warrant a fully automated inspection solution. The catalyst for change is when the inspection room is full of quality inspectors, and it’s time to increase the throughput of quality assessment. The semi-automatic inspection solution allows vials (liquid and lyo), syringes, cartridges and ampoules to be presented to the operator at speed to review before a magnifier, enabling manual rejection to be completed. This allows a single operator to do the job of a room of quality inspectors. This process mimics the standard GAMP manual inspection of the black and white background booths found in manual inspection, allowing for the throughput of high volumes. However, how does the semi-automatic inspection fit into the validation process and allow validation at the relevant levels?

Crush Defects: The Power of Semi-Automatic Inspection

Validation of such semi-automatic inspection must adhere to the United States Pharmacopeia (USP). The following chapters from the USP are regarded as the main literature and source for regulatory information surrounding the visual inspection of injectables:

Chapter 〈790〉 VISIBLE PARTICULATES IN INJECTIONS
Chapter 〈1790〉 VISUAL INSPECTION OF INJECTIONS
Chapter 〈788〉 PARTICULATE MATTER IN INJECTIONS

Crush Defects: The Power of Semi-Automatic Inspection

Chapter 〈790〉 establishes the expectation that each unit of injectable product will be inspected as part of the routine manufacturing process. This inspection should take place at a point when defects are most easily detected; for example, prior to labelling or insertion into a device or combination product. Semi¬automated inspection should only be performed by trained, qualified inspectors. The intent of this inspection is to detect and remove any observed defect. When in doubt, units should be removed

Crush Defects: The Power of Semi-Automatic Inspection

Defect Types

IVS semi-automatic inspection machines with operators allow for the inspection of cosmetic defects in the container and particulate matter within the fluid. The machines will enable the inspection of the following defects:

Cosmetic Defects:
• Glass cracks
• Scratches
• Missing stoppers/caps
• Improper cap/stopper closure
• Glass inclusions

Crush Defects: The Power of Semi-Automatic Inspection

Particulate Matter:
• Fibers
• Glass
• Metal
• Product Related
• Rubber

Crush Defects: The Power of Semi-Automatic Inspection

What are the particle definitions which apply to Semi-Automatic Inspection?

Extrinsic – Highest risk
Particles may originate from many sources. Those that are foreign to the manufacturing process are considered exogenous or “extrinsic” in origin; these include hair, non-process-related fibres, starch, minerals, insect parts, and similar inorganic and organic materials. Extrinsic material is generally a one-time occurrence and should result in the rejection of the affected container in which it is seen; however, elevated levels in the lot may implicate a broader contribution from the same source. These particles may carry an increased risk of microbiological or extractable contamination because less is known about their path before deposition in the product container or their interaction with the product.

Intrinsic – Medium Risk
Other particles are considered “intrinsic”, from within the process. Intrinsic particles may come from processing equipment or primary packaging materials that were either added during processing or not removed during container preparation. These primary product-contact materials may include stainless steel, seals, gaskets, packaging glass and elastomers, fluid transport tubing, and silicone lubricant. Such particles still pose the risk of a foreign body, but generally come from sterile or sanitized materials and more is known about their interactions when in contact with the product.

Inherent- Lower Risk
“Inherent” particles are considered the lowest risk as they are known to be or intended to be associated with specific product formulations. The physical form or nature of inherent particles varies from product to product and includes solutions, suspensions, emulsions, and other drug delivery systems that are designed as particle assemblies (agglomerates, aggregates). Product formulation-related particulate formation should be studied in the development phase and in samples placed on stability to determine the normal characteristics and time-based changes that can occur.

Defects are commonly grouped into classifications based on patient and compliance risk. The most common system uses three groups: critical, major, and minor. Critical defects are those that may cause serious adverse reaction or death of the patient if the product is used. This classification includes any nonconformity that compromises the integrity of the container and thereby risks microbiological contamination of the sterile product. Major defects carry the risk of a temporary impairment or medically reversible reaction, or involve a remote probability of a serious adverse reaction. This classification is also assigned to any defect which causes impairment to the use of the product. These may result in a malfunction that makes the product unusable. Minor defects do not impact product performance or compliance; they are often cosmetic in nature, affecting only product appearance or pharmaceutical elegance.

Crush Defects: The Power of Semi-Automatic Inspection

Inspection criteria

On semi-automatic inspection machines, the vials are spun up at speed prior to reaching the inspection area. This sets any visible particles in motion which aids in detection as stationary particles will be difficult to detect. Upon 100% inspection, visible extrinsic and intrinsic particles should be reliably removed. The test method allows inherent particles to be accepted if the product appearance specification allows inherent particle types. The size of particles reliably detected (2′::70% probability of detection) is generally 150 µm or larger. This Probability of Detection (POD) is dependent on the container characteristics (e.g., size, shape, transparency), inspection conditions (lighting and duration), formulation characteristics (colour and clarity), and particle characteristics (size, shape, colour, and density). For syringes the products are agitated and turned over so that no particulate matter is left in the bung and can be seen.

Crush Defects: The Power of Semi-Automatic Inspection

Critical Inspection Conditions

Light intensity
The results of the inspection process are influenced by the intensity of the light in the inspection zone. In general, increasing the intensity of the light that illuminates the container being inspected will improve inspection performance; 〈790〉 recommends light levels of 2,000-3,750 lux at the point of inspection for routine inspection of clear glass containers. Special attention should be given to assure that inspection is not performed below the lower limit of 2,000 lux.

Background and Contrast
Contrast between the defect of interest and the surrounding background is required for detection, and increased contrast improves detection. The use of both black and white backgrounds is described in 〈790〉, as well as other global pharmacopoeias. The use of both backgrounds provides good contrast for a wide range of particulate and container defects, which can be light or dark in appearance.

Inspection Rate
This is controlled by the roller travel speed. Sufficient time must be provided to allow for thorough inspection of each container; chapter 〈790〉 specifies a reference time of 10 s/container (5s each against both black and white backgrounds).

Magnification
Some inspection processes use a large magnifier to increase image size and thus increase the probability of detecting and rejecting containers with defects near the threshold of detection. Most lenses used during this inspection process are around X2 magnification.

Crush Defects: The Power of Semi-Automatic Inspection

Overall, Semi-Automatic Inspection is a necessary step when pharmaceutical manufacturers move into medium-volume production. Validating such systems allows production to ramp up to the required volume without the need for manual benches and provides capacity for future volumes before fully automated vision inspection with vision systems. More details on how to use Semi-Automatic inspection for regulatory compliance can be found here.

Machine vision: The key to safer, smarter medical devices

Posted on

In this month’s blog, we are discussing the highly regulated field of medical device manufacturing, where precision and compliance are not just desirable—they are mandatory. The growth of the medical devices industry means there have been substantial advances in the manufacturing of more complex devices. The complexity of these products demands high-end technology to ensure each product is of the correct quality for human use, so what’s the key to producing safer, smarter medical devices? Machine vision is one of the most important technologies that allow for the achievement of precision and compliance within the required regulations. In this article, we will touch on the aspect of machine vision in medical device manufacturing and its role in ensuring precision while serving regulatory compliance.

How Machine Vision Works in Manufacturing

Machine vision (often referred to in a factory context as industrial vision or vision systems) uses imaging techniques and software implementation to perform visual inspections on the manufacturing line for many applications where quality control inspection is required at speed. Using cameras, sensors, and complex algorithms—Visual Perception allows machines to develop a sense of sight by capturing images with the camera(s) and processing them in order to enable decision-making from visual data.

When it comes to the medical device manufacturing process, machine vision systems play a vital role in ensuring the highest levels of precision and quality. Examples include systems that can detect the smallest imperfections, measure components with sub-micron precision, and confirm every product is produced to the tightest of tolerances—all automatically.

The Significance of Medical Device Manufacturing Accuracy

Surgical instruments, injectable insulin pens, contact lenses, asthma devices and medical syringes all need high-precision manufacturing, and this is due to several reasons. The devices not only involve patient health and safety but also must be produced at a legal level, to validated specifications. Even tiny damage or a slight deviation from those specifications can lead to the collapse of these products and hence impact human life. The drive towards miniaturisation in medical devices means that more precise parts have to be manufactured than ever before. With devices getting smaller (miniaturisation) and more complex, the margin for error decreases, hence making precision even more critical.

Machine vision systems can easily maintain precision at these levels due to the precise resolution they run at. The vision systems can use advanced imaging techniques, such as 3D vision, to measure dimensions, check alignments, and verify components are manufactured within tolerance. It makes a quality-assurance check on every unit as it comes off the assembly line to prevent defects and recalls later down the line.

Meeting Compliance Regulations

The medical device industry is heavily regulated by organizations like the FDA (Food and Drug Administration) in the United States, European Medicines Agency (EMA), and other national and international bodies that set relevant standards. Regulations govern every dimension of device manufacture, from design and production to testing and distribution.

The Current Good Manufacturing Practice (CGMP) regulations issued by the FDA draw attention to quality requirements needed throughout the manufacturing process. Machine vision systems can automatically inspect components and end products according to specifications, allowing manufacturers to validate to the relevant regulations.

Machine vision systems can produce extremely stringent inspection records with the traceability required to prove compliance against established regulatory standards. This is crucial specifically for audit purposes; if needed, manufacturers must be able to show evidence that their product was manufactured according to regulations.

Using Machine Vision in the Medical Devices Production

Each of these stages using Machine Vision to ensure precision-orientation compliance with the final product. A few of the main uses are:
• Surface Inspection: Vision systems can detect scratches, cracks, and foreign material present on the surface during the manufacturing process, which may lead to a device failure in operation.
• Dimensional Measurement: Vision metrology is used for measuring component dimensions to ensure they meet design specifications.
• Assembly Verification: Machine vision ensures that components are assembled correctly, reducing defect risk from the assembly process.
• Label Verification: Understanding which label goes on which batch is essential for compliance. Machine vision systems can inspect that labels are properly positioned and the writing on them is accurate.
• Packaging Inspection: Proper packaging ensures that devices remain sterile and protected while in transit. Machine vision systems ensure that packaging is not defective and is properly sealed.
• Code Reading: Machine vision systems check and verify datamatrix & barcodes on products and packaging to maintain correct product identification throughout the supply chain.

Optimising Efficiency and Lowering Costs

The machine vision system has more benefits and is likely to improve manufacturing not only in a precise manner but also with strict compliance while increasing the efficiency of manufacturing, resulting in reduced costs. This will help speed up the throughput from these systems, where automated inspection and measurement can be done much faster and more consistently than a team of human inspectors.

Machine vision products also reduce human error, which could cost in relation to medical device manufacturing. Where a simple mistake can cause the manufacturer to recall defective products, leading them to legal lawsuits and damaging their reputation. Catching defects early can help to minimise these risks and costs, which is why machine vision offers such a huge benefit.

Manufacturers can also use inspection data to analyse trends and uncover underlying issues before they lead to defects, allowing for higher uptime with preventative maintenance of the lines.

Future Trends in Medical Device Manufacturing and Machine Vision

As technology advances, the function of machine vision systems will continue to develop in medical device manufacturing. Here are some of the likely trends we can expect to see over time.
• Integration of Artificial Intelligence (AI) with Machine Vision Systems: AI and machine learning are increasingly incorporated into machine vision systems, further advancing image analysis and pattern recognition. This will enable even greater precision and the possibility of detecting less visible imperfections that conventional vision systems might overlook.
• 3D Vision: Where 2D vision systems are traditionally used, the third dimension is being used more. Due to cross-verification, 3D vision systems provide a complete view of the components (compared to a single-plane 2D view), offering better measurement and inspection capabilities over complex shapes and assemblies.
• High-Speed Vision Systems: The push for higher-speed vision systems goes hand in hand with the increased manufacturing speeds being implemented today, and the higher speeds of machines increase demand for accurate inspection.
• Edge Computing: With the advent of edge computing, machine vision systems can process data locally with greater power and accuracy to eliminate latency for real-time decisions. This is crucial in applications that need instant feedback to change the manufacturing process.

Conclusion

Medical device manufacturing can be significantly enhanced with the incorporation of machine vision that provides superior levels of precision and compliance. For manufacturers, it also ensures compliance with rigorous regulatory needs by automating inspections and measurements with machine vision systems. While still advancing, the necessity of machine vision in this industry is only likely to become more critical as time goes on and improvements are made in efficiency, accuracy, and overall manufacturing quality.

How to boost AI machine vision systems

Posted on

Artificial Intelligence (AI) machine vision continues to develop, and using AI deep learning (DL) in automated machine vision inspection has become a valid option in those applications where clear trends, learning, and data are available for process inspection. We continue to see strong growth in the use of AI vision systems. But at what point is a decision made on which machine vision process is best for the application deployment, and when should you boost your machine vision inspection to include an element of 100% AI deep learning inspection? And how do you improve and continue to turbo-charge your AI inspection.

Let’s drill down on what we need to apply AI machine vision. To capture training data, we need a consistent setup with known samples to capture a whole series of images. High-resolution images can capture more details, which can significantly improve the accuracy of the vision system. We need images for training the AI classifier engine, images to test the classifier within the training process, and finally, a set of unseen images to confirm the trained classifier is working as it should be. Some applications are naturally more akin to AI machine vision – such as surface inspection, surface anomalies, or subtle changes to a product’s appearance.

This differs from applications that require specific data to be calculated, such as automated gauging and metrology, where particular parts need to be measured with a vision system to an exact tolerance; this is not an application suited or achievable with AI vision systems (as no such data is available from the AI classifier).

We should also be mindful of how the production systems can be deployed. As AI requires a dataset to train and work with in many cases, the system will need to be installed and images captured from the live system before a determination can be made if AI machine vision is applicable in this case. For example, in medical device and pharmaceutical applications, this is difficult, as the solution needs to be proven and validation paperwork completed before the final installation qualification, and it’s hard to validate an AI model. But in other application areas cameras can be installed and images captured as part of the installation process. Always remember the Pros and Cons of Artificial Intelligence in machine vision.

How can we boost the AI machine vision algorithm?

The first step in boosting the AI machine vision algorithm involves showing new images, more deviations from the normal, and selecting the ability to account for skew, changes in size, and overall deviation of the image. So, it’s important to make the most out of the available data by using data augmentation techniques. This includes rotating, flipping, scaling, and adding noise to the images. Data augmentation increases the diversity of the training data without needing to collect new images, which helps prevent overfitting and improves the generalisation capabilities of the AI model.

It’s important to ensure that your dataset represents the real-world scenarios your machine vision system will encounter. A balanced dataset that includes various angles, lighting conditions, and object variations is crucial for training a robust AI model. Including diverse examples helps avoid biases and makes the system more adaptable and accurate.

It’s important to ensure all elements are handled, so all versions of the reject are known and seen. The classifier will still flag those parts with deviations it has not seen before, but you can boost the AI by providing it with more data to work with. However, there is a trade-off between the time needed to retrain the classifier on a decent GPU PC and the boost you can give to the AI machine vision calculation. It’s best to utilise hardware accelerators like GPUs, TPUs, or FPGAs to speed up the training and inference of the machine vision AI models.

In general, the more data the AI has and the larger the network, the more accurate the analysis will be. Implementing a methodology for continuous integration and deployment (CI/CD) to streamline updating models in production is crucial.

Another way to boost the AI machine vision system is to combine its functionality with traditional machine vision algorithms. We do this often, knowing AI is not a panacea for all application needs. So, use an algorithm, for example, to find specific error states, and then use an AI classification to drill down on ambiguous or hard-to-spot surface/defect deviations.

Continuous Monitoring and Updating of the AI vision system

It’s extremely important to regularly monitor the performance of a machine vision system in production. We generally set up automated systems to track key performance metrics and detect any degradation in performance over time. Continuous monitoring allows timely interventions and updates to maintain the AI model’s high accuracy and reliability. Remember, the environment in which machine vision systems operate can change over time. Periodically update your models with new data to stay accurate and relevant.

Boosting AI machine vision systems involves a holistic approach that includes enhancing data quality, utilising advanced algorithms, ensuring real-time processing capabilities, and implementing robust preprocessing and postprocessing techniques. Additionally, focusing on robustness, generalisation, and continuous monitoring ensures that the system remains accurate and reliable. By following these strategies, you can significantly improve and boost the performance and applicability of your AI machine vision systems, unlocking their full potential across many industry sectors and strengthening their use in industrial automation.

Artificial Intelligence (AI) machine vision

The magic behind machine vision: Clever use of optics and mirrors

Posted on

Sometimes, we’re tasked with some of the most complex machine vision and automated inspection tasks. In addition to our standard off-the-shelf solutions in the medical device, pharmaceutical, and automotive markets, we often get involved with the development of new applications for machine vision inspection. Usually, this consists of adapting our standard solutions to a client’s specific application requirement. A standard machine vision camera or system cannot be applied due to either the complexity of the inspection requirements, the mechanical handling, part surface condition, cycle time needs or simply the limitation to how the part can be presented to the vision system. Something new is needed.

This complex issue is often discussed in customer and engineering meetings, and the IVS engineering team often has to think outside the box and perhaps use some “magic”! This is when the optics (machine vision lenses) need to be used in a unique way to achieve the desired result. This can be done in several ways, typically combining optical components, mirrors, and lighting.

So what optics and lenses are used in machine vision?
The primary function of any lens is to collect light scattered by an object and reproduce a picture of the object on a light-sensitive ‘sensor’ (often CCD or CMOS-based); this is the machine vision camera head. The image captured is then used for high-speed industrial image processing. When selecting optics, several parameters must be considered, including the area to be imaged (field of view), the thickness of the object or features of interest (depth of field), the lens-to-object distance (working distance), the intensity of light, the optics type (e.g. telecentric), and so on.

Therefore, the following fundamental parameters must be evaluated in optics.
Field of View (FoV) is the entire area the lens can see and image on the camera sensor.
Working distance (WD) is the distance between the object and the lens at which the image is in the sharpest focus.
Depth of Field (DoF) is the maximum range over which an object seems to be in acceptable focus.
In addition, the vision system magnification (ratio of sensor size to field-of-view) and resolution (the shortest distance between two points that can still be distinguished as separate points) must also be appraised.

But what if the direct line of the site cannot be seen due to a limitation in the presentation or space available? This is when the magic happens. Optical path planning is used to understand the camera position (or cameras) relative to the part and angle of view required. Combining a set of mirrors in the line of site makes it possible to see over a longer working distance than possible without. The same parameters apply (FoV, WD, DoF etc), but a long optical path can be folded with mirrors to fit into a more compact space. This optical path can be folded multiple times using several mirrors to provide one continuous view through a long path. Cameras can also be mounted closely with several mirrored views to offer a 360-degree view of the product. This is especially true in smaller product inspections like pill or tablet vision inspection checks. Multiple mirrors allow all sides and angles of the pill to be seen, providing a comprehensive automated inspection for surface defects and imperfections for such applications. This technique is also applied in the subtle art of needle tip inspection (in medical syringes) using vision systems, with mirror systems to allow for a complete view of the tip from all angles, sometimes providing a single image with two views of the same product. The optical paths, the effect of specific machine vision lighting and the mechanical design must all be considered to create the most robust solution.

So when our engineers say they will be using “magic” optics, I no longer look for a wand to be waved but know that a very clever application of optics and mirrors will be designed to provide a robust solution for the automated vision system task at hand.

Why industrial vision systems are replacing operators in factory automation (and what they aren’t doing!)

Posted on

This month, we’re following on from last month’s post.
We’ve all seen films showing the dystopian factory landscape a few hundred years in the future. Flying cars against a grey backdrop of industrial scenery. Robots serving customers at a fictitious future bar, and probably – if they showed a production line, it would be full of automatons and robots. We will drill down on the role that industrial vision systems can have on the real factory production line. Is it the case that in the future, all production processes will be replaced with automated assembly, coupled with real-time quality inspection? Intelligent robots which can handle every production process?

Let’s start with what’s happening now, today. It’s certainly the case that machine vision has developed rapidly over the last thirty years. We’ve gone from slow scientific image processing on a 386 computer (look it up!) to AI deep learning running at hundreds of frames per second to allow the real-time computation and assessment of an image in computer vision. But this doesn’t mean that every process has (or can be) replaced with industrial vision systems, but what’s stopping it from being so?

Well, the main barrier to this is the replication of the human dexterity and ability to manoeuvre the product in delicate (compared to an industrial automated process) ways. In tandem with this is the ability for a human to immediately switch to a different product mix, size and type of product. So you could have a human operator construct a simple bearing in a matter of seconds, inspect it – and then ask them to build next a completely different product, such as a small medical device. With some training and show how, this information exchange can be done and the human operator adapts immediately. But a robot would struggle. Not to say it couldn’t be done, but with today’s technology, it takes effort. This is the goal of Industry 5.0, the next step generation from Industry 4.0 flexible manufacturing. Industry 5.0 is a human-centric approach, so industry understanding that humans working in tandem with robotics, vision, and automation is the best approach.

So we need to look at industrial vision in the context of not just whether the image processing can be completed (be it with traditional algorithms or AI deep learning), but also how we can present the product meaningfully at the correct angles to assess it. Usually, these vision inspection checks are part and parcel with an assembly operation, which requires specific handling and movement (which a human is good at). This is the reason most adoption of machine vision happens on high throughput, low variation manufacturing lines – such as large-scale medical device production where the same product is validated and will be manufactured in the same way over many years. This makes sense – the payback is there and automation can be applied for assembly and inspection in one.

But what are the drivers for replacing people on production lines with automated vision inspection?
If the task can be done with automation and vision inspection, it makes sense to do it. Vision inspection is highly accurate, works at speed and is easy to maintain. Then there are the comparisons comparing the processes to a human operator. Vision systems don’t take breaks, don’t get tired, and robots don’t go to parties (I’ve never seen a robot at a party) – so they don’t start work in the morning with their minds not on the task at hand! So, it makes sense to move towards automated machine vision inspection wherever possible in a production process, and this represents huge growth in the adoption of industrial vision systems in the coming years.

If the job is highly complex in terms of intricate build, with many parts and variants, then robots, automation and vision systems are not so easy to deploy. However, with the promise of Industry 5.0, we have the template of the future – moving towards an appreciation of including the human operator in factory automation – combining humans with robots, automation, augmented reality, AI and automated inspection. So, the dystopian future might not be as bad as the filmmakers make out, with human operators still being an integral part of the production process.

Catalysts of change: How robots, machine vision and AI are changing the automotive manufacturing landscape

Posted on

This month, we are discussing the technological and cultural shift in the automotive manufacturing environment caused by the advent of robotics and machine vision, coupled with the development of AI vision systems. We also drill down on the drivers of change and how augmented reality will shape the future of manufacturing technology for automotive production.

Robots in Manufacturing
Robots in manufacturing have been used for over sixty years, with the first commercial robots used in mass production in the 1960s – these were leviathans with large, unwieldy pneumatic arms, but they paved the way for what was to come in the 1970s. At this time, it was estimated that the USA had 200 such robots [1] used in manufacturing; by 1980, this was 4,000 – and now there are estimated to be more than 3 million robots in operation [2]. During this time the machine vision industry has grown to provide the “eyes” for the robot. Machine vision uses camera sensor technology to capture images from the environment for analysis to confirm either location (when it comes to robot feedback), quality assessment (from presence verification through to gauging checks) or simply for photo capture for warranty protection.

The automotive industry is a large user of mass-production robots and vision systems. This is primarily due to the overall size of the end unit (i.e. a built car), along with vision systems for confirmation of quality due to the acceptable parts per million failure rate (which is extremely low!). Robots allow repetitive and precise assembly tasks to be completed accurately every time, reducing the need for manual labour and providing a faster speed for manufacturing.

Automating with industrial robots is one of the most effective ways to reduce automotive manufacturing expenses. Factory robots help reduce labour, material, and utility expenses. Robotic automation reduces human involvement in manufacturing, lowering wages, benefits, and worker injury claims.

AI Machine Vision Systems
Deep learning in the context of industrial machine vision teaches robots and machines to do what comes naturally to humans, i.e. to learn by example. New multi-layered “bio-inspired” deep neural networks allow the latest machine vision solutions to mimic the human brain activity in learning a task, thus allowing vision systems to recognise images, perceive trends and understand subtle changes in images that represent defects. [3]

Machine vision performs well at quantitatively measuring a highly structured scene with a consistent camera resolution, optics and lighting. Deep learning can handle defect variations that require an understanding of the tolerable deviations from the control medium, for example, where there are changes in texture, lighting, shading or distortion in the image. Deep-learning vision systems can be used in surface inspection, object recognition, component detection and part identification. AI deep learning helps in situations where traditional machine vision may struggle, such as parts with varying size, shape, contrast and brightness due to production and process constraints.

Augmented Reality in Production Environments
In industrial manufacturing, machine vision is primarily concerned with quality control, which is the automatic visual identification of a component, product, or subassembly to ensure that it is proper. This can refer to measurement, the presence of an object, reading a code, or verifying a print. Combining augmented, mixed reality with automated machine vision operations creates a platform for increased efficiency. There is a shift towards utilising AI machine vision systems (where applicable!) to improve the reliability of some quality control checks in vision systems. Then, combine that assessment with the operator.

Consider an operator or assembly worker sitting in front of a workstation wearing wearable technology like the HoloLens or Apple Vision Pro (an augmented reality headset). An operator could be putting together a sophisticated unit with several pieces. They can see the tangible objects around them, including components and assemblies. They can still interact with digital content, such as a shared document that updates in real time to the cloud or assembly instructions. That is essentially the promise of mixed reality.

The mixed-reality device uses animated prompts, 3D projections, and instructions to guide the operator through the process. A machine vision camera is situated above the operator, providing a view of the scene below. As each step is completed and a part is assembled, the vision system automatically inspects the product. Pass and fail criteria can be automatically projected into the operator’s field of sight in the mixed reality environment, allowing them to continue building while knowing the part has been inspected. In the event of a rejected part, the operator receives a new sequence “beamed” into their projection, containing instructions on how to proceed with the failed assembly and where to place it. When integrated with machine vision inspection, mixed reality has become a standard aspect of the production process. Data, statistics, and important quality information from the machine vision system are shown in real time in the operator’s field of view.

Robots, AI Vision Systems and Augmented Reality
Let’s fast forward a few years and see what the future looks like. Well, it’s a combination of all these technologies as they continue to develop and mature. So AI vision systems mounted on cobots help with the manual assembly, while the operator wears an augmented reality headset to direct and guide the process for an unskilled worker. Workers are tracked while all tasks are being quality assessed and confirmed, while data is stored for traceability and warranty protection.

In direct automotive manufacturing, vision systems will become easier to use as large AI deep-learning datasets become more available for specific quality control tasks. These datasets will continue to evolve as more automotive and tier one and two suppliers use AI vision to monitor their production, allowing quicker deployment and high-accuracy quality control assessment across the factory floor.


References

Why you should transition your quality inspection from shadowgraphs to automated vision inspection

Posted on
Download article

The shadowgraph used to be the preferred method, or perhaps the only method for the quality control department to check the adherence to measurement data. These typically vertical projectors have a plate bed, with parts laid on it, and the light path is vertical. This allows a silhouette and profile of the part to be shown to the operators as an exact representation of the part in black, but projected at a much higher magnification. These projectors are best for flat, flexible parts, washers, O rings, gaskets and parts that are too small to hold, such as very small shafts, screws etc. They are also used in orthopedic manufacturing to view profiles and edges of joint parts.

So, this old-fashioned method involves projecting a silhouette of the edge of the product or component. This projection is then overlaid with a clear (black-lined) drawing of the product showing some additional areas where the component shouldn’t stray into. The operator chooses the overlay from a bunch available, based on the part number of the part. This is laid over the projection and manually lined up with the tolerances shown as a thick band in key areas for quality checking. For example, if you’re an orthopedic joint manufacturer and are coating your product with an additive material, you might use a shadowgraph to check the edges of the new layer. The shadowgraph will project a magnified representation of the component to compare against.

Some optical comparators and shadowgraphs have developed to a level where they are combined with a more modern graphical user interface (GUI) and are digital to make overlays easier to check, coupled with the ability to take manual measurements.

Ultimately though, all these types of manual shadowgraphs take time to use, are manually intensive and are totally reliant on an operator to make the ultimate decision on what is a pass and a fail. You also have no record of what the projection looked like and the whole process takes time. This all leads to a loss of productivity and makes your Six Sigma quality levels dependent on an operator’s quality check, and as we all know, manual inspection (even 200% manual inspection) is not a reliable method for manufacturers to keep their quality levels in check. Operators get tired and there is a lack of consistency between different people in the quality assessment team.

But for those products with critical-to-quality measurements, the old shadowgraph methodology can now be replaced with a fully automated metrology vision system to reduce the reliance on the operator in making decisions, coupled with the ability to save SPC data and information.

In this case, the integration of ultra-high-resolution vision systems, in tandem with cutting-edge optics and collimated lighting, presents a formidable alternative to manual shadowgraphs, ushering in an era of automation. With this advancement, operators can load the product, press a button, and swiftly obtain automatic results confirming adherence to specified tolerances. Furthermore, all pertinent data can be meticulously recorded, with the option to enable serial number tracking, ensuring the preservation of individual product photos for future warranty claims. This transition not only illuminates manufacturing processes but elevates them to a brighter, six-sigma enhanced quality realm. Moreover, these systems can be rigorously validated to GAMP/ISPE standards, furnishing customers with an unwavering assurance of consistently superior quality outcomes.

For your upcoming quality initiatives or product launches, consider pivoting towards automated vision system machines, transcending the limitations of antiquated shadowgraph technology and operator-dependent assessments. Embrace the superior approach that is now within reach, and steer your manufacturing quality towards a future defined by precision, efficiency, and reliability.

By embracing automated vision systems, manufacturers streamline processes and ensure unparalleled precision and consistency in product quality. This shift represents a pivotal leap forward in manufacturing excellence, empowering businesses to meet and exceed customer quality expectations while navigating the ever-evolving landscape of modern production.

How to create data-driven manufacturing using automated vision metrology systems

Posted on
Download article

Applying automated vision metrology technologies for process control before, during and after assembly and machining/moulding is now a prerequisite in any production environment, especially in the orthopaedic and additive manufacturing industries. This month, we’re drilling down on the data and knowledge you can create for your manufacturing process when deploying vision metrology systems for production quality control. Let’s face it – we all want to know where we are versus the production schedule!

What are “vision metrology” systems and machines?

It’s the use of non-contact vision systems and machines using metrology grade optics, lighting and software algorithms to allow increased throughput of quality control inspection and data collection in real-time for the production process.

What are the benefits of vision metrology systems?

Before we look at the data available on such systems and how they can be utilised to create data-driven production decisions, we should look at the benefits of automated vision metrology systems. This, combined with the data output, clearly focuses on why such systems should be deployed in today’s production environments.

Increased throughput. One of the main drivers and benefits of using automated vision metrology systems is the higher throughput compared to older contact and probing CMM approaches. Coupled with integration into autoload and autotending options, the higher production rates of deploying automated vision metrology allow for faster production and increased output from the factory door.

Better yields. As the systems are non-contact, there is no chance of damage or marking of the product, which is the risk with probing and contact inspection solutions. Checking quality through automated vision allows yields to improve across the production process.

Faster reaction to manufacturing issues. Data is king in the fast-moving production environment. Real-time defect detection, seeing spikes in quality and identifying quality issues quicker, allows production managers to react faster to changes or faults in manufacturing faults. Vision metrology allows immediate review and analysis of statistical process control on a batch, shift and ongoing basis. Monitor live progress of work orders as they happen.

Less downtime. With no contact points and fewer moving parts, maintenance downtime is minimal when applying automated vision metrology machines. Maintenance can be used for other tasks and increases productivity across the manufacturing floor.

Guarantees your quality level. Ultimately, the automated vision metrology machine is a goal-keeper. This allows the quality and production team to sleep easy, knowing that parts are not only being inspected, but also providing warranty protection via a photo save of every product through production.

How do we get data from an automated vision metrology quality control system?

This is where the use of such vision technology gets interesting. The central aspect is the quality control and automated checking of crucial measurements (usually critical to quality CTQ characteristics). Still, the data driving the quality decision is paramount for the manager. The overall data covers a range of data objects that could potentially be needed when running a modern manufacturing plant. These include OEE data, SPC information, shift records, KPI metrics, continuous improvement information, root cause analysis, and data visualisation.

OEE Data (Overall Equipment Effectiveness). The simplest way to calculate OEE is the ratio of fully productive time to planned production time. Fully productive time is just another way of saying manufacturing only good parts as quickly as possible (at the ideal cycle time) with no stop time. A complete OEE analysis includes availability (run time/planned production time), performance ((ideal cycle time x total count)/run time) and quality (good count/total count). The vision metrology data set allows this data to be easily calculated. The data is stored within the automated vision metrology device or sent directly to the factory information system.

Statistical Process Control (SPC) is used in industrial manufacturing quality control to manage, monitor, and maintain production processes. The formal term is “the use of statistical techniques to control a process or production method”. The idea is to make a process as efficient as possible while producing products within conformance specifications with as little scrap as possible. Vision metrology allows vital characteristics to be analysed at speed on 100% of product, compared to the slow, manual load CMM route. SPC data can be displayed on the vision system HMI, and immediate decisions can be made on the process performance and data presentation.

Shift records are part of the validation and access control system for the vision system used in vision metrology. This stops operators from accessing certain features and logging all information against a specific shift operator or team. Data analytics allow trends in operator handling and contribute to the data provided by the system. You may see a spike in quality concerns from a specific shift or operator; these trends can easily be tracked and traced.

KPI metrics are Key Performance Indicators normally specific to the manufacturing site. However, these invariably include monitoring the quality statistics, OEE, process parameters, environmental factors (e.g. temperature and humidity), shift data, downtime and metrology measurements. This is all part of understanding how certain situations impact the output from processes and how levels can be maintained.

Continuous Improvement is the process to improve a manufacturing facility’s product and service quality. Therefore, the data and images from any automated vision metrology machine play an important role in providing live data related to Six Sigma and real-time monitoring of the ppm (parts per million) failure rate within the facility.

Root cause analysis of manufacturing issues is challenging without tangible data from a quality inspection source. The ability to review product images at the point of inspection (with the date and time stamped information), coupled with the specific measurement data, allows easier and quicker root cause analysis when problems crop up.

Data visualisation is now built into modern automated vision metrology systems. The production manager and operators can see quality concerns, spikes in rejects, counts of good and bad, shift data, and live SPC all in a single screen. This can be displayed locally or sent immediately to the factory information system.

By utilising the latest-generation automated vision metrology systems manufacturers can now build a fully-integrated data driven production environment. They are allowing easier control based on accurate, immediate information and photos of products running through production.

The ultimate guide to Unique Device Identification (UDI) directives for medical devices (and how to comply!)

Posted on

This month, we’re talking about the print you see on medical devices and in-vitro products used for the traceability and serialisation of the product – called the UDI or Unique Device Identification mark. In the ever-evolving world of healthcare, new niche manufacturers and big medical device industrial corporations find themselves at a crossroads. Change has arrived within the Medical Device Regulation, and many large and small companies have yet to strike the right chord, grappling with the choices and steps essential to orchestrate compliance to the latest specifications. Automated vision systems are required at every stage of the production and packaging stages to confirm compliance and adherence to the standards for UDI. Vision systems ensure quality and tracking and provide a visible record (through photo save) of every product stage during production, safeguarding medical devices, orthopaedic and in-vitro product producers.

UDI

What is UDI?

The UDI system identifies medical devices uniformly and consistently throughout their distribution and use by healthcare providers and consumers. Most medical devices are required to have a UDI on their label and packaging, as well as on the product itself for select devices.

Each medical device must have an identification code: the UDI is a one-of-a-kind code that serves as the “access key” to the product information stored on EUDAMED. The UDI code can be numeric or alphanumeric and is divided into two parts:

DI (Device Identifier): A unique, fixed code (maximum of 25 digits) that identifies a device’s version or model and its maker.

Production Identifier (PI): It is a variable code associated with the device’s production data, such as batch number, expiry date, manufacturing date, etc. UDI-DI codes are supplied by authorised agencies such as GS1, HIBCC, ICCBBA, or IFA and can be freely selected by the MD maker.

UDI

UDIs have been phased in over time, beginning with the most dangerous equipment, such as heart valves and pacemakers. So, what’s the latest directive on this?

Navigating the Terrain of MDR and IVDR: Anticipating Challenges

The significant challenges presented by the European Medical Devices Regulation (MDR) 2017/745 have come to the fore since its enactment on May 26th, 2021. This regulatory overhaul supersedes the older EU directives, namely MDD and AIMDD. A pivotal aspect of the MDR is its stringent oversight of medical device manufacturing, compelling the display of a unique code for traceability throughout the supply chain – this allows full traceability through the process from the manufacture of the medical device to the final consumer use.

Adding to the regulatory landscape, the In-Vitro Diagnostic Regulation (IVDR) 2017/746 debuted on May 26th, 2022. This regulation seamlessly replaced the previous In-Vitro Diagnostic Medical Devices (IVDMD) regulation, further shaping the intricate framework governing in-vitro diagnostics. The concurrent implementation of MDR and IVDR ushers in a new era of compliance and adaptability for medical devices and diagnostics stakeholders.

What are the different classes of UDI?

Risk Profile Notify or Self-Assessment Medical Devices In-Vitro Diagnostic Medical Devices
High Risk to Low Risk Notified Body Approval Required Pacemakers, Heart Valves, Implanted cerebral simulators Class III Class D Hepatitis B blood-donor screening, ABO blood grouping
Condoms, Lung ventilators, Bone fixation plate Class IIb Class C Blood glucose self-testing, PSA screening, HLA typing
Dental fillings, Surgical clamps, Tracheomotoy Tubes Class IIa Class B Pregnancy self-testing, urine test strips, cholesterol self-testing
Self-Assessment Wheelchairs, spectacles, stethoscopes Class I Class A Clinical chemical analysers, specimen receptacles, prepared selective culture media

Deadlines for the classification to come into force

Implementing the new laws substantially influences many medical devices and in-vitro device companies, which is why the specified compliance dates have been prioritised based on the risk class to which the devices belong. Medical device manufacturers must adopt automated inspection into their processes to track and confirm that the codes are on their products in time for the impending deadlines.

The Act recognises four types of medical devices and four categories of in-vitro devices, which are classified in ascending order based on the degree of risk they provide to public health or the patient.

2023 2025 2027
UDI marking on DM Class III Class I
Direct UDI marking on reusable DM Class III Class II Class I
UDI marking on in-vitro devices (IVD) Class D Class C & B Class A

The term “unique” does not imply that each medical device must be serialised individually but must display a reference to the product placed on the market. In fact, unlike in the pharmaceutical industry, serialisation of devices in the EU is only required for active implantable devices such as pacemakers and defibrillators.

EUDAMED is the European Commission’s IT system for implementing Regulation (EU) 2017/745 on medical devices (MDR) and Regulation (EU) 2017/746 on in-vitro diagnostic medical devices (IVDR). The EUDAMED system contains a central database for medical and in-vitro devices, which is made up of six modules.

EUDAMED is currently used on a voluntary basis for modules that have been made available. The connection to EUDAMED must take place within six months of the release of all platform modules that are fully functioning.

Data can be exchanged with EUDAMED in three methods, depending on the amount of data to be loaded and the required level of automation.

The “visible” format of the UDI is the UDI Carrier, which must contain legible characters in both HRI and AIDC formats. The UDI must be displayed on the device’s label, primary packaging, and all upper packaging layers representing a sealable unit.

These rules mean medical device manufacturers need reliable, traceable automated vision inspection to provide the track and trace capability for automatic aggregation and confirmation of the UDI process.

The UDI is a critical element of the medical device manufacturing process, and therefore, applying vision systems with automated Optical Character Recognition (OCR) and Optical Character Verification (OCV) in conjunction with Print Quality Inspection (PQI) is required for manufacturers to guarantee that they comply with the legislation.

But what are the differences between Optical Character Recognition (OCR) vs Optical Character Verification (OCV) vs Print Quality Inspection (PQI)?
We get asked this often, and how you apply the vision system technology is critical for traceability. To comply with UDI legislation, medical device manufacturers must check that their products comply with the requirements and track them through the production process. Whether you use OCR, OCV, or PQI depends on the manufacturer stage and where you are in the manufacturing process.

UDI

Optical Character Recognition (OCR)

Optical character recognition is based on the need to identify a character from the image presented, effectively recognise the character, and output a string based on the characters “read”. The vision system has no prior knowledge of the string. It therefore mimics the human behaviour of reading from left to right (or any other orientation) the sequence it sees.
You can find more information on OCR here.

Optical Character Verification (OCV)
Optical character verification differs from recognition as the vision system has been told in advance what string it should expect to “read”, and usually, with a class of characters it should expect in a given location. Most UDI vision systems will utilise OCV, as opposed to OCR, as the master database, line PLC (Programmable Logic Controller), cell PLC and laser/label printer should all know what will be marked. Therefore, the UDI vision system checks that the characters marked are verified and readable in that correct location (and have been marked). This then allows for the traceability through the process through nodes of OCV systems throughout the production line.
You can find more information on OCV here.

Print Quality Inspection (PQI)

Print quality inspection methods differ considerably from the identification methods discussed so far. The idea is straightforward: an ideal template of the print to be verified is stored; this template does not necessarily have to be taken from a physically existing ‘‘masterpiece’’, it could also be computer-generated. Next, an image containing all the differences between the ideal template and the current test piece is created. Simply, this can be done by subtracting the reference image from the current image. But applying this method directly in a real-world context will show very quickly that it will always detect considerable deviations between the reference template and test piece, regardless of the actual print quality. This is a consequence of inevitable position variations and image capture and transfer inaccuracies. A change in position by a single pixel will result in conspicuous print defects along every edge of the image. So, print quality inspection is looking for the difference between the print trained and the print found. Therefore, PQI will pick up missing parts of characters, breaks in characters, streaks, lines across the pack and general gross defects of the print.

You can find more information on PQI here.

UDI
UDI

What about Artificial Intelligence (AI) with OCR/OCV for UDI?

Where we’ve been discussing OCR and OCV, we’ve assumed that the training class is based either on a standard OCR font (a unique sans-serif font created in the early days of vision inspection to provide a consistent font, recognisable by humans and vision systems), or a trainable, repeatable fixed font. OCR with AI is a comprehensive deep-learning-based OCR (or OCV) method. This innovative technology advances machine vision towards human reading. AI OCR can localise characters far more robustly than conventional algorithms, regardless of orientation, font type, or polarity. The capacity to automatically arrange characters enables the recognition of entire words. This significantly improves recognition performance because misinterpreting characters with similar appearances is avoided. Large images can be handled more robustly with this technique, and the AI OCR result offers a list of character candidates with matching confidence ratings, which can be utilised to improve the recognition results further. Users benefit from enhanced overall stability and the opportunity to address a broader range of potential applications due to expanded character support. Sometimes, this approach can be used. However, it should be noted that it can make validation difficult due to the need for a data set prior to learning, compared to a traditional setup.

Do all my UDI vision inspection systems require validation?

Yes, all the vision systems for UDI inspection must be validated to GAMP/ISPE levels. Having the systems in place is one thing, but the formal testing and validation paperwork is needed to comply fully with the legislation. You can find more information on validation in one of our blog articles here.

UDI
UDI

Industrial vision systems allow for the 100% inspection of medical devices at high production speeds, providing a final gatekeeper to prevent any rogue product from exiting the factory gates. These systems, in combination with factory information communication and track & trace capability, allow the complete tracking of products to UDI standards through the factory process.

Machine vision vs computer vision vs image processing

Posted on

We’re often asked – what’s the difference between machine vision and computer vision, and how do they both compare with image processing? It’s a valid question as the two seemingly similar fields of machine vision and computer vision both relate to the interpretation of digital images to produce a result from the image processing of photos or images.

What is machine vision?
Machine vision has become a key technology in the area of quality control and inspection in manufacturing. This is due to increasing quality demands and the improvements it offers over and above human vision and manual operation.

Definition
It is the ability of a machine to consequentially sense an object, capture an image of that object, and then process and analyse the information for decision-making.

In essence, machine vision enables manufacturing equipment to ‘see’ and ‘think’. It is an exciting field that combines technologies and processes to automate complex or mundane visual inspection tasks and precisely guide manufacturing activity. In an industrial context, it is often referred to as ‘Industrial Vision’ or ‘Vision Systems’. As the raw image in machine vision are generally captured by a camera connected to the system in “real-time” (compared to computer vision), the disciplines of physics relating to optical technology, lighting and filtering are also part of the understanding in the machine vision world.

So, how does this compare to computer vision and image processing?
There is a great deal of overlap between the various disciplines, but there are clear distinctions between the groups.

Input Output
Image processing Image is processed using algorithms to correct, edit or process an image to create a new better image. Enhanced image is returned.
Computer vision Image/video is analysed using algorithms in often uncontrollable/unpredictable circumstances. Image understanding, prediction & learning to inform actions such as segmentation, recognition & reconstruction.
Machine vision Use of camera/video to analyse images in industrial settings under more predictable circumstances. Image understanding & learning to inform manufacturing processes.

Image processing has its roots in neurobiology and scientific analysis. This was primarily down to the limitations of processor speeds when image processing came into the fore in the 1970’s and 1980’s—for example, processing single images following the capture from a microscope-mounted camera.

When processors became quicker and algorithms for image processing became more adept at high throughput analysis, image processing moved into the industrial environment, and industrial line control was added to the mix. For the first time, this allowed an analysis of components, parts and assemblies as part of an automated assembly line to be quantified, checked and routed dependent on the quality assessment, so machine vision (the “eyes of the production line”) was born. Unsurprisingly, the first industry to adopt this technology en mass was the PCB and semiconductor manufacturers, as the enormous boom in electronics took place during the 1980s.

Computer vision crosses the two disciplines, creating a Venn diagram of overlap between the three areas, as shown below.

Machine vision vs computer vision vs image processing

As you can see, the work on artificial intelligence relating to deep learning has its roots in computer vision. Still, over the last few years, this has moved over into machine vision, as the image processing learnt from biological vision, coupled with cognitive vision moves slowly into the area of general purpose machine vision & vision systems. It’s rarely a two-way street, but some machine vision research moves back into computer vision and general image processing worlds. Given the more extensive reach of the computer vision industry, compared to the more niche machine vision industry, new algorithms and developments are driven by the broader computer vision industry.

Machine vision, computer vision and image processing are now broadly intertwined, and the developments across each sector have a knock-on effect on the growth of new algorithms in each discipline.

IVS helps global contact lens manufacturers achieve error-free lenses (at speed!)

Posted on

This month, we are talking about automated visual inspection of cosmetic defects in contact lens inspection, a major area of expertise in the IVS portfolio. Contact lens production falls between medical device manufacturing and pharmaceuticals manufacturing, but it does mean that it falls within the GAMP/ISPE/FDA validation requirements when it comes to manufacturing quality and line validation.

The use of vision technology for the inspection of contact lenses is intricate. Due to the wide variety of lens profiles, powers, and types (such as spherical, toric, aspherical etc.), manufacturers historically relied on human operators to visually verify each lens before it was shipped to a customer. Cosmetic defects in lenses range from scratches and inclusions, through to edge defects, mould polymerisation problems and overall deformity in the contact lens.

The Challenge: Manual Checks are Prone to Human-Error and are Slow!

The global contact lens industry must maintain the highest standards of quality due to the specialist nature of its products. Errors or faults in contact lenses can lead to negative recalls, financial losses, or a drop in brand loyalty and integrity – all issues that could otherwise be very difficult to rectify. However, manually checking lenses has become extremely time-consuming, inefficient, and error-prone. Due to the brain’s tendency to rectify errors naturally, human error is a common cause of oversights. This makes it next to impossible to spot faults by hand, therefore many errors are missed, and contact lens defects can be passed to the consumer.

Coupled with this, many global contact lens manufacturing brands run their production at high speed, which means using human inspectors simply becomes untenable based on the volume and speed of production; so in the last twenty-five years, contact lens manufacturers have moved to high-speed, high-accuracy automated visual inspection, negating the need for manual checks.

To ensure their products’ quality, accuracy, and consistency, many of the world’s top contact lens manufacturing brands have turned to IVS to solve their automated inspection and verification problems. This can be either wet or dry contact lens inspection, depending on the specific nature of the production process.

The Solution: Minimize Errors and Ensure Quality Through IVS Automated Visual Inspection

IVS contact lens inspection solutions have been developed over many years based on a deep and long heritage in contact lens inspection and the required FDA/GAMP validation. The vision system drills down on the specific contact lens defects, allowing finite, highly accurate quality checks and inspection coupled with the latest generation AI vision inspection techniques. This allows the maximum yield achievable, coupled with full reporting and statistical analysis, to give the data feedback to the factory information system, providing real-time, reliable information to the contact lens manufacturer. Failed lenses are rejected, allowing 100% quality inspected contact lenses to be packed and shipped.

IVS have inspection systems for manual load, automated and high-speed contact lens inspection, allowing our solutions to cover the full range of contact lens manufacturers – from small start-ups to the major high-level production volume producers. Whatever the size of your contact lens production, we have validated solutions to FDA/ISPE/GAMP for automated contact lens inspection.

Contact lens manufacturers now have no requirement for manual visual inspection by operators; this can all be achieved through validated automated visual inspection of cosmetic defects, providing high-yield, high-throughput quality inspection results.

More details: https://www.industrialvision.co.uk/wp-content/uploads/2023/05/IVS-Lens-Inspection-Solutions.pdf

4 common red flags in your product design that will cause you trouble during automated vision inspection

Posted on

This month, we’re going to discuss the aspects of design in your product which could affect the ability to apply automated visual inspection using vision systems to the production process. This can be a frustrating aspect for the engineering and quality managers when they’ve been asked to safeguard bad products going out of the door using machine vision, but there has been no attempt up the chain at the product design stages to either design the product in an effective way to make it simple to manufacture or to apply vision inspection to the process. Once the design has been signed off and validated, it’s often extremely difficult to get small changes instigated, and engineering is then tasked with the mass production and assembly of a product, with in-line visual inspection required.

Very often in medical device manufacturing, due to the long lifetime and cycles of products, the design was conceived months, or even years before it goes through a process of small batch production, before moving to mass production at high run rates – all this with the underlining aspect of validation to GAMP and FDA production protocols. This is the point when automated inspection will be integral to the quality and safeguarding of the consumer as part of the manufacturing process.

From our experience, we see the following red flags which could have been addressed at the design phase, which make the vision system more difficult to design and run.

1. Variants. Often, there can be too many variants of a product with additional variants of sub-assemblies, which means the number of final product variations can run to hundreds. While the vision system may be able to cope with this variation, it makes set-up more costly and ongoing maintenance more difficult. Sometimes, these variations are due to multiple customers down the line wanting subtle changes or unique part numbers, which the design team happily accommodate with no thought on the impact on the manufacturing process. The more variation which can be removed from the design phase, the easier and more cost-effective a solution for machine vision inspection will be.

2. Lack of specification in surface/material conditions. The settings for a machine vision system will depend on a product’s surface and material conditions. This is especially critical in surface inspection or any process where the segmentation of the background is required from the foreground – so presence verification, inclusion detection and even edge detection for gauging and measurement in machine vision. Suppose conditions vary too much due to a lack of specification on colour or surface being included in the design. This variation can make the vision system more susceptible to false failures or higher maintenance. While latest-generation artificial intelligence (AI) vision systems are making such analysis easier and less prone to failure in these conditions, having as little deviation in your incoming image quality for vision inspection makes sense. The design should include a specification on the surface conditions expected for repeatable manufacturing and inspection, which, for example, in plastic mould colour specification can be challenging to specify.

3. Inability to see the inspection area. In the past, we’ve been asked to automatically inspect surfaces which can’t be seen due to a lack of clear view from any angle. The product’s design is such that perhaps datum edges and points are too far away and not easily accessible. The design could be such that an overhang or other feature obscures the view. This is often the case in large sub-assemblies where the datum edge can be in a completely different area and angle to the area where vision inspection will be applied. Design engineers should assess the feasibility of applying automated inspection to the process to account for how easily the vision system can access certain areas and conditions. We often discuss such applications with manufacturing engineers, with the frustration from their side that the design engineers could have incorporated cut-outs or areas which would have provided a clear line of sight for the vision system to be applied.

4. Handling. Products should be designed to be easily handled, fed and presented to a vision system as part of the process. There will nearly always be a contact surface on tooling for handling, which will then limit the area where the vision system can view. Sometimes, this surface requires critical inspection, and the site is obscured due to the product’s design. For example, in syringe body inspection, the plastic body can be held by the top and the bottom and rotated, meaning some areas of the mould are not visible. If a lip had been designed in, the part could have been inspected twice, held differently, thus providing 100% inspection of the entire surface. Small design changes can impact both the manufacturing and automated vision inspection process. Smaller products can be inspected on glass dials and fixtures to mitigate this issue, but the design team should consider these issues at the start of the design process.

Product design needs to get input from the manufacturing and quality team early in the design process. Now vision systems are a standard part of the modern automated production process and an invaluable tool in the industry 4.0 flexible production concept; thought needs to be given to how subtle design changes and more detailed specification requirements on colour and surface quality in the design stage can help the integration and robust use of vision systems.