Industrial Computer Vision
AI for Quality Assurance
Transform your quality assurance processes with AI-based automation
Manufacturing and producing products, equipment, and components requires optimization between speed, precision, and scale. Computer Vision technologies are a powerful tool to automate repetitive, high-accuracy inspection processes without jeopardizing the quality requirements of the process. Our AI and software technology competencies enable you to establish automated, standardized, and reliable quality assurance processes that can be easily adjusted according to changing business requirements.
Welding quality inspections
In the automotive industry, manufacturers need to conduct rigorous testing activities as part of their quality control processes. This testing is often a time and human resource-consuming activity. In one of the recent customer cases, we focused on automating the inspection of welding seams, reducing the time spent on quality assurance for each part. The development and deployed solution allowed improvements in both productivity and quality. The solution is used in several factories located on different continents.
Electronics assembly inspection automation
In electronics manufacturing, computer vision can be used to automate different visual inspection tasks. In the Through Hole Technology (THT) process, automated inspection can be established after the manual assembly to detect potential defects of components, whether they are missing, lifted, or wrong components, or if they are broken or have a wrong orientation. Our technology is designed to support high-mix, low-volume production setup and support the fast introduction of new products and product variations to the process.
Label check automation
One of the final stages in manufacturing processes is placing product, batch, or pallet labels on product packages. Although the systems placing the labels are usually automated, this process may result in defects. By establishing a computer vision-based label check automation process, all kinds of potential label problems can be avoided whether they are non-existing, non-readable, wrong position, wrong content, or missing key data. The identified defects are stored in a database, visualized to operators, and a fault message is sent instantly to production line automation for corrective actions.
Cable quality inspection
In cable manufacturing, AI-based inspection automation can be established for different phases of cable production. One of the recent customer cases was related to inspecting the outer sheath of the cable. The cables are moving at medium-to-high speeds during production, requiring precise data collection that also covers 360 degrees of the cable. The developed computer vision methodology enabled the reliable detection of sheath bumps, scratches, droplets, sudden changes in sheath width, and incorrect color changes
Benefits
Significant cost and time savings through reliable inspection automation
Improved quality through a more standard quality assurance process
Reduced risks and liabilities through improved quality
Improved compliance and traceability
AI for Factory Automation
Streamlining your business operations through AI-based automation
Industrial factories and sites require extensive usage of cameras, sensors, and other monitoring equipment to follow up the state of production lines and working environment. Computer vision technologies are a natural tool to enable cost-efficient 24/7 monitoring automation with the ability to instantly respond to changes in monitored processes or equipment.
Conveyor Belt Monitoring
Critical production lines and conveyor belts need to be monitored to secure continuous, optimal performance according to production requirements. By utilizing advanced Computer Vision technologies, the monitoring can be easily automated to detect risks that require e.g. belt maintenance or, in case of a critical event, to stop the production line. The solutions are many times carrying out multiple tasks simultaneously that can include measuring and dimension checking, material identification, object detection, and segmentation, to name a few.
Computer vision-based process optimization
Computer vision-based process optimization can be established to a wide range of use cases. Some of the typical ones include monitoring the amount of end product in a package to avoid over-quality or following a production material quality against a desired criteria. In both these use cases a camera-based data collection is established to provide a constant data feed of the process to AI-based analysis, which is trained to understand the set production criteria. As an outcome of the optimization, the operator is given real-time information on the process status. The application can be complemented to give recommendations to which direction the process should be directed.
Automated counting
The ability to count objects as part of process follow-up, monitoring, or reporting is a task that has to be supported and implemented in a trustworthy and traceable manner. At the same time, it is often difficult to automate. The challenges may relate to a highly dynamic line of activity, which was the case in one of our recent use cases within the entertainment sector. In other cases, poor visibility and complex counting conditions present obstacles, such as in one of our construction-related use cases in Japan. Our generic competencies in advanced detection, classification, segmentation, and specialized counting methods enabled us to effectively implement reliable counting automation applications to our customers.
Optimizing operations and maintenance process
Infrastructure maintenance can be optimized by combining different as-built data sources, like CAD files or 3D models, to up-to-date image data. We supported our customer in infrastructure maintenance by developing an application that enabled comparing images taken on the site with mobile phones to as-built documentation to validate that the right kind of components would be installed to infrastructures requiring maintenance. The application included object detection and measurement functions, enabling the component verification to be done in a reliable and scalable manner.
Benefits
Significant cost savings through a higher level of automation
A standardized 24/7 process monitoring
Better transparency and traceability of the processes
Optimal usage of professionals’ time – from monitoring and operative actions
Safety with AI
Improving Safety through Human Analytics
Securing safety is among the highest priorities in industrial work sites. Safety procedures and infrastructure include visual surveillance for monitoring work tasks and identifying potential safety risks. We have developed a wide range of computer vision-based human analytics applications to support multiple use cases, ranging from movement analysis and crowd density analysis to human posture detection. In addition, the applications have been complemented with action initiation capabilities to enable instant response when needed.
Safety monitoring at factory sites
Safety is of the highest importance for manufacturing companies, and at the same time, a difficult challenge to cope with. Our computer vision solutions enable monitoring and detecting body movements from a distance under challenging conditions while meeting privacy requirements. The solutions may involve monitoring human and equipment movement simultaneously, enabling instant alerts in case of identified safety risks. As a result, our customers have achieved higher safety in their operative sites.
Safety equipment and human pose tracking
Two very common human analytics tasks that have a wide range of use cases are monitoring the use of safety equipment as well as tracking human poses and movement of a person. Safety equipment tracking is easy to implement, which enables any company to establish such an important monitoring process at a very low cost. Human pose estimation use cases relate typically to situations when there is a risk of abnormal behavior and authorities are responsible for reaching such situations instantly. We have established a solid foundation to provide these services to our customers.
Benefits
Reduced safety & security risks through AI
Cost-efficient and reliable 24/7 surveillance
Wide range of Human Analytics use cases implemented
High flexibility to meet requirements of new use cases