Predictive Maintenance Software

by Rod Dunne on July 9, 2010

in Articles, Performance

  • Share
  • Share

Predictive maintenance software was originally developed in the airline and auto-industry for statistically calculating the lifespan of components, and therefore when they should be replaced. This article looks at the background to maintenance software and the current usage trends (in particular, in preventative maintenance planning of high-end computer repair jobs).

The development of predictive maintenance software was originally conceived to use statistical analysis to establish the lifespan of components in machines – if the failure point can be estimated then the component can be replaced as part of planned maintenance, and before the error occurs.

Establishing the life span for a component can incorporate factors such as the running time (effective usage time), the environment in which it is used (low/high strain) and the treatment that the component receives (is it pushed to its maximum intended workload?). Maintenance software goes one step further by using historical data of failures along with statistical analysis to establish a working lifespan to within a specific range of error.

Once the lifespan can be measured, or at least estimated, then the software has got to notify engineers/administrators that predictive maintenance is due on the component. Most systems automate the production of work orders/tickets for engineers to include as part of planned maintenance schedules (i.e. the variety of computer or PC repair jobs).

The real advantage of using these tools is that they limit the amount of component failures that occur along with the resulting downtime that this incurs. For the organization, work streams and work orders can be automated and more easily planned (so predictive maintenance schedules can feed into staffing needs, overtime requirements and inventory control).

One perceived disadvantage that some feel predictive maintenance software introduces is the wastage of working components before they have completed their entire lifespan. It is fair to say that these disposal costs may be weighty (especially if a component could exceed the estimated life expectancy) but this needs to be balanced against the costs of failure.

A failing component may result in systems downtime, backlogs and staff redeployment costs. In truth, each industry has to decide between reactive maintenance (where you fix error messages after they occur) and using predictive/preventative maintenance software based on their own unique requirements.

The current usage trend is to use preventative maintenance software to automate repairing the predicted failures. One notable usage of this is in high-end computer systems (million dollar server systems for banks, stock exchanges or web server farms). These computer systems are often built with redundant components pre-installed (RAM, motherboards, CPU, etc.) that the maintenance software can turn on during scheduled computer repair and maintenance for rerouting processing, when it feels that some components are reaching the end of their lifespan.

Related posts:
- Why is my computer so slow all of a sudden? Solutions to rapidly worsening performance issues.

Previous post:

Next post: