View original at www.manufacturing.net

Kay Jenkins How do you think doctors know that high blood pressure can lead to critical health issues and conditions? And how did the medical field come up with the numbers that distinguish high blood pressure from normal blood pressure readings? The answer: medical science kept records over time on these readings and the outcomes from them, and performed analytics on the data that revealed the association of high blood pressure to deadly outcomes. The medical field thus became able to predict and prevent or treat certain related health conditions based on a data point. The same is possible with manufacturing assets. Most manufacturing companies have already moved toward a preventive maintenance approach, with scheduled maintenance at set intervals, usually determined by time. These schedules often begin with tribal knowledge, gleaned from years of hands-on experience. But excessive maintenance can be a drain on resources, since 30 percent of preventive maintenance activities are carried out too frequently. Moving to more advanced maintenance strategies may seem like a huge leap for small to mid-size manufacturers. However, to reduce the intimidation factor, consider maintenance as a continuum of gradual transitions based on information from the machines. Listening to the Machines Research by IBM found that 89 percent of asset failures occur at random, and those are difficult to prevent with planned maintenance. Condition-based maintenance can provide an appropriate strategy for these random failures. Asset conditions trigger maintenance with enough lead time before failure, so work can be completed before performance falls below tolerance. To begin moving toward this usage-driven model of preventive maintenance, it is critical to have data. Identify the appropriate operational metric for your production. It is important that this data be obtained from the assets. Ideally sensors would be integrated with the EAM, although operator input is sufficient. […]