Problem-free equipment operation is critical to maintaining optimum production yield and quality for common metalworking processes such as stamping and welding. If something is wrong with the equipment, or if the equipment unexpectedly fails, a large amount of output may have to be scrapped, or production shut down. In either case, the manufacturer will have to bear a significant loss.
To address these issues, Good Tech Instruments combines its machine learning algorithm with ADLINK's MCM-100 edge inference device to build a smart production equipment monitoring platform. The MCM-100 enables metal processors to use artificial intelligence (AI) and Industrial Internet of Things (IIoT) implementations for real-time monitoring on all types of metal processing equipment to enhance production performance.
According to Vincent Hsu, Technical Manager at Good Tech Instruments, metal processing lines use a wide range of different equipment categorized as continuous or discontinuous based on operational use. Continuous machines like motors operate around the clock while discontinuous machines include large robotic arms, stamping presses and automatic welders. These machines are valuable tools to manufacturers, and any problem or failure can result in a large economic loss.
Hsu takes automobile manufacturing as an example. Stamping is a common process in the manufacturing of automotive parts. A defect in a mold will influence output quality, but such an issue may not be discovered until a quality check is performed later in the workflow. However, by then, a large quantity of substandard items may already have been produced. Aside from financial losses, there is also time wasted when the production line has to be halted for equipment inspection and repair. These problems do not just plague the automobile manufacturing industry, but can happen to the production process of any metal parts, from screws to components for aerospace applications.
A diverse range of smart equipment monitoring systems is available on the market to address such issues. These systems all make use of IIoT sensor networks and AI to detect and analyze equipment conditions to warn factory managers before equipment failure occurs. AI machine learning algorithms play an important role in these systems. Hsu points out that in a discontinuous metalworking process, sensors can be used to detect equipment conditions including vibration of stamping machines, and electric current variation of welding machines, which produce repeated, predictable patterns. In conventional manufacturing environments, machine operators develop a sense about such patterns from long-term experience. Smart manufacturing systems turn workers' experience into machine learning algorithms that can be used to warn factory managers when anomalies in vibration or electric current patterns are detected.
AI machine learning algorithms require a large amount of data for training. As such, most of the systems on the market using machine learning algorithms need to first deploy sensors to gather data. Then, when equipment anomalies happen, the sensor data is marked to train the machine to learn from the data. This approach is undesirable in that manufacturers often have to spend a lot of time and money on data collection. However, Good Tech Instruments has developed an equipment monitoring system that can be up and running in an instant. According to Hsu, Good Tech Instruments preloads a trained vibration or electric current model on the end device, saving the manufacturer from having to collect the data. Then, the platform quickly learns from the range of data specified by the manufacturer. The system that has completed the training is capable of automatic tracking and detection, logging the operating conditions of every piece of equipment and reporting the data back to the manager, who can then adjust the equipment to optimize operation.
ADLINK's MCM-100 plays an instrumental role enabling the ready-to-use design of Good Tech Instruments' equipment monitoring system. Hsu notes that conventional equipment monitoring systems do not come with end devices. Instead, manufacturers have to additionally purchase industrial computers and extraction cards and assemble them into a complete system after testing and calibration. This not only extends the development cycle but also puts extra burden on their engineers. The MCM-100 eliminates such hassles. Designed specifically for smart manufacturing, the MCM-100 is a ready-to-use vibration/condition monitoring platform for rotating machinery that features high-performance edge computing. With built-in four-channel 24-bit 128 kS/s simultaneous sampling of analog input, it can monitor signals sent from multiple sensors at the same time. It also supports a wide range of I/O interfaces for optimal interoperability. Leveraging the MCM-100's high level of integration and powerful performance, Good Tech Instruments can focus efforts on system development, allowing the system to be up and running very quickly.
Hsu comments that "smartization" is the most important trend in manufacturing in recent years and equipment monitoring is regarded by most manufacturers as the first step to the introduction of smart manufacturing systems. With smartization being a new concept to the manufacturing industry, it takes tremendous resources for manufacturers to introduce and calibrate a smart manufacturing system. Delivering optimal performance through a high level of software and hardware integration, the AI-based equipment monitor platform built by Good Tech Instruments and ADLINK enables quick introduction at significantly reduced costs so that manufacturers can begin to enjoy the benefits of smart manufacturing sooner, and have a smooth start embarking on digital transformation.
Good Tech Instruments' VMS-ML machine learning system enables real-time monitoring for stamping and punching processes
According to Vincent Hsu, technical manager, Good Tech Instruments, equipment monitoring can significantly boost production line performance
DIGITIMES' editorial team was not involved in the creation or production of this content. Companies looking to contribute commercial news or press releases are welcome to contact us.