Hitachi Develops Fault Prediction Technology for Production Equipment
Integration Enhances Hi-PHMA Technology, Optimizing Production with Data-Driven Insights
“What … surprised me was that MATLAB … [provided] support at these different steps. The tools I’ve [used] here are just some of the tools I use very often. I’ve used them countless times, and I really trust them.”
Key Outcomes
- MATLAB enabled cost-effective data visualization, accelerating the development of fault prediction technology.
- Cross-platform code generation capabilities in MATLAB improved efficiency and reduced time to market for the final product.
- MATLAB toolboxes enabled the extraction of domain features from data generated in normal and abnormal operating modes.
In automated manufacturing industries, fault prediction technology is crucial for optimizing production sites that utilize numerous motors of varying sizes. The efficiency of these motor systems underpins fault prediction capabilities.
Hitachi, a global R&D leader that focuses on industrial automation, developed motor sensing technology for free data acquisition and high-precision monitoring, along with motion recognition technology for servo motors and a Simulink® model for fault simulation. It implemented an edge section for online monitoring and used portable devices for offline fault detection and predictive maintenance.
The team developed an IoT network centered around sensor technology, using MATLAB® for data-based modeling to gather extensive data from production sites. MATLAB facilitated algorithm development, verification, and deployment. Data on three-phase alternating current during capacitor malfunctions enabled effective motor fault detection.
Hitachi integrated the MATLAB algorithm with a system developed in a third-party programming language. This integration provided flexible online and offline solutions, enhancing the Hi-PHMA industrial equipment fault prediction technology.
