What is autonomous machine vision?
19 March 2019
Yonatan Hyatt, CTO of Inspekto, a company at the forefront of Autonomous Machine Vision explains the fundamentals of this emerging category
Automation and robotics are now commonplace across manufacturing environments. According to the International Federation of Robotics (IFR), there will be a projected 630,000 industrial robots shipped to customers in the year 2021. However, automation has not been so quick to catch on in the quality assurance (QA) industry.
According to the World Quality Report 2018-9, automation is the biggest bottleneck holding back QA and testing today. The report also found that artificial intelligence is becoming more important in testing, though 51 per cent of respondents have experienced difficulty integrating AI with their existing applications.
Another main message of the report was that QA professionals must acquire additional skills to keep up with new technologies in the market, such as mastering artificial intelligence and blockchain. Since the launch of Autonomous Machine Vision at VISION in Stuttgart in November 2018, all this has changed for visual QA for gating and sorting in manufacturing.
Autonomous Machine Vision gives QA managers and plant managers complete control over their QA processes. It requires little cost, effort and time to install and run. Importantly, rather than handing the reins of a machine vision project over to an external systems integrator to design and build a custom system for a particular point on the production line, manufacturers can do it in house.
In under an hour, the plant’s own personnel can install an INSPEKTO S70 and begin assessing the quality of products on their production line. Because it is as simple as Plug and InspectTM, the manufacturer can have complete independence and control over QA in the facility.
Based on sophisticated artificial intelligence technology, Autonomous Machine Vision not only means that no external systems integrator is needed, but also that the plant’s own staff require no additional skills or training to set it up or operate it. The system’s algorithm can optimise the camera and illumination settings for the object and environment and then detect and locate the object without any input from the operator. The system only requires a limited number of good sample references and the system learns the properties of a gold standard product.
The system is ready to go, straight out of the box ─ the QA manager simply has to draw a polygon to mark their areas of interest on the object to be inspected. Once in operation, the system will compare each image with the gold standard, verifying both the shape tolerances and surface variations to identify any defects.
QA is everywhere
If the manufacturer then decides that the system would be best placed at another point on the production line, it can be moved as and when required and set up in just a few minutes. Alternatively, the manufacturer can purchase multiple systems to install at every required point on the line. This is because the system can be purchased at one tenth of the cost of a traditional solution and will adapt to each new product and environment in minutes.
Autonomous Machine Vision therefore enables Total QA – visual quality assurance at every step of the production line. The manufacturer can identify a defective product at the exact stage that the defect was introduced. This improves yield and prevents scrap, by ensuring no further time or energy is wasted on a defective part and that it is not combined with a good part later to form a faulty product.
Total QA also enables the manufacturer to optimise their line. By performing root cause analysis using data that represents the full picture of what is happening, the QA manager can trace a defect back to its source. They can then take action to replace or perform maintenance on the equipment at fault and prevent defects from being introduced in the future. In this way, the manufacturer can optimise and streamline their plant, protecting their customers from defective products and protecting their facility from scrap and flawed manufacturing processes.
Autonomous Machine Vision addresses the QA manager’s true needs. Unlike traditional machine vision solutions, suited only to the integrator’s needs, Autonomous Machine Vision gives the QA manager complete control of where, when and how visual QA systems are deployed.
- Autonomous Machine Vision gives QA managers and plant managers complete control over their QA processes
- A manufacturer can identify a defective product at the exact stage that the defect was introduced
- Algorithms optimise the camera and illumination settings and then detect and locate the object without any operator input