When companies fail to conduct regular maintenance work for production jigs, it can result in unnecessary equipment shutdown or a delayed response to jig abnormalities. Using AI detection of abnormalities and notify the user of maintenance needs can reduce maintenance hours by up to 67%. What is the secret to its success?
Jigs and fixtures need replacing before they deteriorate otherwise workpiece precision is at risk. Conventionally, operators decide when to conduct maintenance based on the test result. However, this can result in wasteful work, such as stopping the equipment for jig replacement, only to find out that the jig had no problem. Or because a machine error occurred due to a delayed response to jig abnormalities.
AI detection with no experience needed
With the introduction of Mitsubishi Electric’s data science tool MELSOFT MaiLab, AI can diagnose the deterioration of the jig in real time without having to stop the equipment. It notifies the user of the optimum maintenance time, leading to significant reduction in maintenance work hours. In this example, before the analysis monthly maintenance time averaged 28.5 hours. With the introduction of MaiLab, this fell to 9.5 hours, or a reduction of 67%.
Automatic creation of AI models
The automated Machine Learning (AutoML) function automatically creates an AI detection model for diagnosis based on past test results. If an abnormality occurs, the real-time diagnostics automatically send a notification to the operator. Their action prevents machine malfunctions that could have happened without maintenance.
Importantly, users can easily introduce this system without having any AI expertise! The outcome is a high accuracy, real-time diagnosis that previously needed data scientists.
Eight-month payback
In this application, the cost was £14k including system configuration and a personal computer. Development and implementation took six months, and the outcome was a 67% reduction in maintenance time. This represented an eight-month payback period when including reduced labour cost with lower maintenance frequency and the profit increase.
System overview
This system consists of a personal computer with the Data Science Tool MELSOFT MaiLab installed and an existing data storage server. It is easy to set up just by connecting the data server and the test equipment to MELSOFT MaiLab.
After the system introduction, the user configures the setting in a conversational form on MELSOFT MaiLab. It then automatically creates an AI detection model that meets the selected objective, to allow real-time abnormality detection diagnosis.
To learn more about how we can assist you, please use the contact form below.
