Pleora’s Visual Inspection System Ensures End-to-End Quality for DICA Electronics
DICA Electronics Ltd is deploying Pleora’s Visual Inspection System to reduce manufacturing quality escapes and gather key data from manual processes to help speed root cause analysis. The system uniquely requires just one image to start using AI, with continuous and transparent training based on operator actions to improve and speed automated decision support. With just one good image, Inspection apps for incoming, in-process, and final quality control steps automatically compare products to a “golden reference” and visually highlight differences and deviations for an operator. As the operator accepts or rejects potential errors, the AI model is transparently trained based on their decisions. After even just one inspection, the AI model will start automatically suggesting a decision for the operator. Over time, the speed and accuracy of automated decision-making will improve as the system continuously learns from operator preferences. In comparison, most AI inspection tools require numerous good and bad images plus time-consuming and expensive algorithm development before they can be deployed in production.
Visual Anomaly Detection: Opportunities and Challenges
Clarifai is pleased to announce pre-GA product offering of PatchCore-based visual anomaly detection model, as part of our visual inspection solution package for manufacturing which also consists of various purpose-built visual detection and segmentation models, custom workflows and reference application templates.
Users only need a few hundred images of normal examples for training, and ~10 anomalous examples for each category for calibration & testing only, especially with more homogeneous background and more focused region-of-interest.
An implementation of YOLO-family algorithms in classifying the product quality for the acrylonitrile butadiene styrene metallization
In the traditional electroplating industry of Acrylonitrile Butadiene Styrene (ABS), quality control inspection of the product surface is usually performed with the naked eye. However, these defects on the surface of electroplated products are minor and easily ignored under reflective conditions. If the number of defectiveness and samples is too large, manual inspection will be challenging and time-consuming. We innovatively applied additive manufacturing (AM) to design and assemble an automatic optical inspection (AOI) system with the latest progress of artificial intelligence. The system can identify defects on the reflective surface of the plated product. Based on the deep learning framework from You Only Look Once (YOLO), we successfully started the neural network model on graphics processing unit (GPU) using the family of YOLO algorithms: from v2 to v5. Finally, our efforts showed an accuracy rate over an average of 70 percentage for detecting real-time video data in production lines. We also compare the classification performance among various YOLO algorithms. Our visual inspection efforts significantly reduce the labor cost of visual inspection in the electroplating industry and show its vision in smart manufacturing.
Industry 4.0 and the pursuit of resiliency
There are two parts to the Zero D story. Visual inspection and asset performance management (APM). Visual inspection uses computer vision models focused on quality inspection. APM uses machine learning models based on time series data to determine health of assets and probable failures in the future. Toyota is using Maximo Visual Inspection, and now they are also using the Maximo Asset Performance Management (APM) suite. They tested Maximo APM on some of their machinery that does liquid cooling and found that was another problem area for them. By implementing the software into this pilot, they are now able to monitor the asset health 24×7 and predict probability of failure in the future.
Why AI software companies are betting on small data to spot manufacturing defects
The deep-learning algorithms that have come to dominate many of the technologies consumers and businesspeople interact with today are trained and improved by ingesting huge quantities of data. But because product defects show up so rarely, most manufacturers don’t have millions, thousands or even hundreds of examples of a particular type of flaw they need to watch out for. In some cases, they might only have 20 or 30 photos of a windshield chip or small pipe fracture, for example.
Because labeling inconsistencies can trip up deep-learning models, Landing AI aims to alleviate the confusion. The company’s software has features that help isolate inconsistencies and assist teams of inspectors in coming to agreement on taxonomy. “The inconsistencies in labels are pervasive,” said Ng. “A lot of these problems are fundamentally ambiguous.”
Applying Artificial Intelligence to Food Tray Production
Neurala, a supplier of AI (artificial intelligence)-based visual inspection technology, began working with apetito to detect cases of the five most reported missing components from meal trays using Neurala’s Vision Inspection Automation (VIA) software. VIA consists of two software programs, Inspector and Brain Builder. Using these programs, apetito was able to build anomaly-detecting systems in 10-20 minutes and immediately begin testing.
With apetito’s earlier weight-based inspection system, the company could only flag an incomplete tray, without understanding what was missing. With VIA’s ability to inspect multiple regions of interest on the trays, apetito can now see specifically which components are missing and identify trends in missing components to avoid their occurence in the future.
Ford presents its prestigious IT Innovation Award to IBM
The Maximo Visual Inspection platform can help reduce defects and downtime, as well as enable quick action and issue resolution. Ford deployed the solution at several plants and embedded it into multiple inspection points per plant. The goal was to help detect and correct automobile body defects during the production process. These defects are often hard to spot and represent risks to customer satisfaction.
Although computer vision for quality has been around for 30 years, the lightweight and portable nature of our solution — which is based on a standard iPhone and makes use of readily available hardware — really got Ford’s attention. Any of their employees can use the solution, anywhere, even while objects are in motion.
Ford found the system easy to train and deploy, without needing data scientists. The system learned quickly from images of acceptable and defective work, so it was up and running within weeks, and the implementation costs were lower than most alternatives. The ability to deliver AI-enabled automation using an intuitive process, in their plants, with approachable technology, will allow Ford to scale out rapidly to other facilities. Ford immediately saw measurable success in the reduction of defects.