August 9, 2016 | Written by: Bill Timme
Categorized: Defence & Intelligence
As a former submarine commanding officer and later as the US Navy admiral responsible for all submarine engineering, maintenance and certifications, I can assure you that I was a firm believer in making sure equipment worked correctly. There was no margin for error, and the goal was always to make sure that the number of times a submarine surfaced equaled the number of times it dove.
That being said, it was very frustrating to invest time, resources and attention to the disassembly and inspection of equipment that appeared to be working perfectly well only to find out – it actually was just fine! After the routine calendar driven inspection, the complex piece of equipment would be put back together in a confined space by someone who only has basic skills instead of a master craftsman in the perfect environment of a factory. Sometimes the reassembly created more problems that it resolved with errors, lack of correct tools, rush to complete, introduction of contaminants and a host of other issues. Real readiness failures were created by the very maintenance that was supposed to ensure the system, vehicle, ship, etc. is ready to go.
A way to address those concerns is Preventive Maintenance and Quality (PMQ). This is an analytics effort to identify equipment that is trending toward failure and performing corrective maintenance before catastrophic failure at, as Murphy’s Law would predict, the most inconvenient time. Instrumentation, failure modeling and analysis are instrumental to success. A simple example to illustrate is taking sound cuts of a pump motor. Over time and with experience, rising noise levels of a bearing could indicate a trend toward failure. Replacing the bearing when operations permit allows expenditure of resources exactly where needed – the other correctly functioning bearings are left alone. If this is extended to many systems you can see that needless expenditures would go down, reliability would go up resulting in a much more effective use of scare funding and resources. In the US alone $80B (with a B) of maintenance costs are wasted in improper, unnecessary or incorrect.
IBM has partnered with the University of South Carolina’s College of Engineering and Computing Condition Based Maintenance Center of Excellence to provide new insights into predictive maintenance. Les Eisner at minute 40 at discusses practical application of predictive maintenance, asset management and Internet of Things at the conference Interconnect 2015.