posted on 1/25/2011 5:48:33 PM
The mVision technology includes manufacturing intelligence, the MIMOSA data model, and enterprise integration for a holistic approach to predictive equipment maintenance and management.
Asset performance management provider Mtelligence is delivering a holistic approach to equipment condition monitoring by combining manufacturing intelligence, open standards, and pre-built enterprise integration adapters to asset management and computerized maintenance management systems.
Delivered in a software-as-a-service model, the recently released technology, called mVision, is said to be a machine learning platform that incorporates neural network technology, statistical processing, and signal processing algorithms to determine predictors of equipment failure.
According to Mtelligence, manufacturers spend a large portion of their budget on the maintenance of plant automation systems, yet are still plagued by equipment failure, wasteful energy usage, and reduced product quality. mVision applies a combination of machine learning techniques, including sensor connections between equipment and transactional data coming from EAM/CMMS systems, a library of intelligent processing filters, and integration with the full suite of Mtelligence asset performance software, to act on performance problems.
“Machine learning hasn't hit mainstream in manufacturing, in part due to the effort required to build an accurate model and get the data needed to train the system,” said Alex Bates, Mtelligence chief technology officer, in a statement.
To solve that problem, Mtelligence engaged a team of data mining experts to develop mVision, thereby helping plant engineers and maintenance managers make sense of the mountains of data coming from various software packages, Bates said.
The mVision platform includes pre-built adapters for maintenance and automation systems, converting all data into the MIMOSA (Machinery Information Management Open Systems Alliance) open standard for modeling asset, maintenance, and condition monitoring data.
The use of the MIMOSA open standard data model and messaging protocol enables mVision to integrate with a wide variety of data sources, and also enables extensibility. For example, production line health indicators from mVision can be integrated into MES (manufacturing execution systems) and ERP packages for optimal scheduling decisions based on capability forecast, the company said.
“In continuous manufacturing environments, one of the toughest decisions is when to stop production for preventive maintenance,” Bates said. “Preventive maintenance is often deferred, leading to increased component failures and reactive maintenance. With mVision, maintenance is prioritized based on the actual condition of the equipment.”
mVision includes pre-built adapters for SAP, IBM Maximo, Infor EAM, Infor Hansen, Ventyx EMPAC, JD Edwards, and others. On the operations side, drivers are provided for plant historians including OSIsoft PI System, Wonderware Historian, GE Proficy, Honeywell PHD, and numerous other automation packages.
By offering a pay-as-you-go SaaS-based approach, Mtelligence enables manufacturers to take advantage of recent advances in the field of machine learning and neural networks without a large upfront capital investment. The solution supports both on-premises deployment and off-site hosting in Microsoft’s Azure cloud platform.