Kolektor (KOL) is known as the world's largest manufacturer of commutators, the second largest European manufacturer of slip rings and the second largest European manufacturer of plastic-bonded magnetic products. Injection moulding is enabling technology for production of all these parts. The scope of this pilot project will be a production line where Kolektor produces one type of products. The aim of this pilot is to detect, possibly predict, and remove the cause of the process failure as soon as possible, ideally in real-time.
In the process of real-time process monitoring, we are planning to:
(i) Collect moulding process parameters;
(ii) Monitor environmental parameters;
(iii) Inspect moulding tool and moulded parts;
(iv) Introduce AQ control loops at the operational level.
Based on the collected data and by applying control loops, advanced analytics and artificial intelligence methods we will better understand the moulding process and will be able to detect anomalies and failures as soon as possible. Because of the geometry of a moulding tool and number of cavities, it is not possible to inspect all cavities at once. Therefore, we are planning to use robots to perform complex moves required for inspection. We would like to study if it is possible to automate the removal of root cause of the bad parts being produced (like cleaning the cavity with dry ice).
| Country: | SI |
Kolektor's Qu4lity project is addressing the real-time injection moulding process monitoring-control. The scope of the pilot project is a production line where Kolektor produces one type of product. The aim of this pilot is to detect, possibly predict, and remove the cause of the process failure as soon as possible, ideally in real-time. Based on the collected data and by applying the control loops, advanced analytics, and artificial intelligence methods we are trying to better understand the moulding process, with the emphasis on detecting anomalies and failures as soon as possible.
We are developing Sinapro IIoT MES/MOM cloud solution (part of the Kolektor Digital Platform) as the cornerstone of the MOM system which enables real-time collecting, evaluating, validating, filtering, checking, and storing of production data. The captured production data can be processed in real-time for the purpose of obtaining various production information, which enables immediate action. MOM function for production analyses with depth learning technology of AI gives users additional and high-quality information’s for fast decisions to achieve zero-defect goals in production.
The Kolektor Digital Platform enables human involvement on various levels. Human operators can monitor, view and inspect created datasets. During the process of model training, the operator can monitor the current state and detailed information of the training process. The Kolektor Digital Platform opens a channel between a data scientist and a decision-making individual in the production line. It is desirable to have multiple people, each assigned to a specific task. The whole process could be split into subtasks - acquiring images on the production line, human expert labeling the images (classification, anomaly,..), data scientist training the model on the new dataset and at the end evaluation of the model and pushing the new (improved) model to the production line.
The Kolektor Digital Platform enables us to automatically collect the data from the shop floor. The Sinapro IIoT enables the connectivity of the pilot production line machines and related IoT devices for real-time production data acquisition and monitoring. The acquired data is afterward used in the off-line machine learning pipelines to produce machine vision predictive models to detect visual injection moulding defects. A pipeline for deploying such off-line machine learning to a HPC cluster is being developed at JSI within the scope of the Kolektor Pilot.
The acquired data is used in on-line prediction of defects. The predicted defects are used to adapt the visual quality inspection with an in-hand camera with a robot. The robot is guided to and between predetermined viewpoints associated with the predicted defects. The robot motion is generated autonomously on-line.
- No items published
With the help of skilled production line workers, the data in the AI platform can be annotated and herewith produce the predictive models for ZDM autonomous quality inspection. The platform gives users the ability to monitor the AQ process (Autonomous Quality) and provide feedback for the ZDM.
To acquire quality data, all involved users and managers must understand some basic data science principles. Machine vision in modern times relies on large amount of consistent data. Data acquisition process begins with organized collection of samples, which should become an integral part of every standardized manufacturing process that involves automated quality inspection or ZDM.