Online Robotics Trade Magazine Industrial Automation, Robots and Driverless Vehicles

If you don't have a RoboticsTomorrow partner account, please register-it's free!

Autonomous Machine Vision (AMV) is a new quality assurance method, which aims to be not only automatic but also autonomous in the process from determining the ideal number of samples required by the system to learning the characteristics of items.

For traditional machine vision solutions, since reflections on the surface can distort the image and cause inaccurate defect detection, inspecting shiny surfaces can be challenging. Because this problem is common in plastics, it usually prevents plastic manufacturers from using automated quality assurance (QA) methods. Yonatan Hyatt, CTO of Inspekto, discusses how autonomous machine vision systems can help. 

Machine vision systems are the backbone of quality assurance and are usually used to check the results of all plastic processing methods, including injection molding, rotational molding, extrusion, compression, etc. These systems can use cameras that analyze products from different angles to identify products with manufacturing defects. The data is then processed by integrated software that detects defective products, and then removed from the production line manually or automatically.      

However, using traditional machine vision systems to inspect plastics has several disadvantages. A common problem is to check highly reflective surfaces. Light reflection will produce certain visual noise, and the system's camera will treat it as a physical feature of the inspected part. The system may in turn classify these characteristics as defects, resulting in false rejections. Similarly, reflections can also cause certain areas of the inspected product to appear hidden, preventing inspections. This means that for plastic packaging or other shiny items, the inability of the system to deal with reflections can be a serious pain point.

Most importantly, the fact that the same production line can create items of different colors and shapes is problematic for traditional machine vision solutions because it can only inspect one product at a time. This is especially common in plastic injection molding, where molds are frequently changed to manufacture various products. 

In order to solve this problem, human workers need to perform quality assurance alone or together with machine vision systems, carefully inspect each product and eliminate defective products. This process can be detailed and repetitive for humans, leading to errors. Defective products may be sent to the company and further rejected, leading to interruption of the logistics chain and expanding waste. 

Autonomous machine vision (AMV) is a new quality assurance method, its design is not only automatic, but also autonomous, from determining the ideal number of samples required by the system to learn the characteristics of the item to automatically adjusting the camera and lighting settings to be inspected The best image of the product. The system's algorithm automatically optimizes the camera and lighting settings to take the best possible image of the object for inspection, and then detects and locates the object without any input from the operator.

INSPEKTO S70, the only AMV system currently on the market, has just been re-launched on the market, with new and improved settings, allowing even the brightest objects to take clear, informative images so that they can be checked reliably. S70 Gen.2 is equipped with 5000K LEDs, which are controlled by pulse width modulation and arranged in several different segments. This allows the system to autonomously control the lighting direction and take multiple images with different light directions and intensities. The images are then merged to create a single non-reflective HDR image. This patent-pending technology makes this system ideal for testing highly reflective plastics. The innovative lighting system is designed to avoid any flicker or light changes seen by the human eye, so nearby staff will not be disturbed by sudden light changes. 

Because AMV systems are flexible and applicable to a wide variety of use cases, they can also allow multiple products to be inspected at the same location on the production line, which is contrary to the capabilities of traditional systems. This makes them very suitable for applications where various items are manufactured on the same production line, such as plastic injection molding.

At the core of the Inspekto system are three independent and coordinated AI-based engines that drive a proprietary electro-optical system. This technology is called Autonomous Machine Vision AI (AMV-AI). The first engine is responsible for real-time adjustment of the operating parameters of the optoelectronic system. Even under constantly changing environmental conditions, it can capture clear and informative images of objects for inspection. The second is the detection and alignment AI module, which allows the system to recognize the parts it sees in the real-time video stream from the camera and determine the best moment to acquire the image to perform the last task-inspection. Finally, the inspection AI engine inspects the parts by comparing the parts with the sample images memorized during the setup, without any expert or AI training in the process.

Because the system is very cost-effective, manufacturers can choose to install multiple inspection systems on the production line to ensure comprehensive quality assurance of plastic products and optimize production. This means that defects can be found early, not just at the end of the production line, so defective products can be removed before more resources are wasted to complete the defective products. Over time, this quality method allows factories to significantly reduce production waste and save energy originally used to produce defective products. 

Autonomous machine vision also allows production plants to reduce downtime. It can take weeks or even months to install a traditional machine vision system and the software it requires. The AMV system can be set up in 45 minutes, and on average, it takes 20 to 30 OK (good) sample items, and no NOK (defective) sample items. 

There is no need to program, train or create rules and collect defective parts to train the system to understand the appearance of these parts. In addition, the AMV system continues to improve over time by collecting more and more data about the products they inspect. 

Using AMV to test the ease of use and cost-effectiveness of plastic products may become a driving force for manufacturers to implement the most advanced technology to avoid wasting materials and help improve manufacturing sustainability. 

There are no comments for this article. Be the first to comment below.

You must log in before you can post a comment. Log in now.

Machine Vision of the Future: How 3D Scanning in Motion Revolutionizes Automation

Opening of the new maxon electromechanical production center

The true flexibility of the robotic system

3D vision system provides network for correct tuna

Offline robot programming for polishing aircraft engine lips

DSI Military Vehicle System Summit

Subscribe| Search| Promote Products| Introduce Your Company| Publish News| List Events

Article| News| Story| Products| Company| Events

© 2010-2021 RoboticsTomorrow-All rights reserved Powered by BTH Management