The Queuing Theory in Breakdown Maintenance Planning
Lizile Xulu
Asset Performance Management | Certified Six Sigma Coach | Certified Reliability ABC Practitioner
In maintenance we dread facing machine breakdowns as they are often connected to the stigma of bad maintenance practices and is often the number one reason for operations to claim loss of production.
A frequent method of maintenance in several organisations that do not have a robust maintenance setup is to engage in breakdown maintenance.
Although breakdown maintenance is sometimes a deliberate or conscious strategy adopted by smaller organisations that lack the budget and sophistication of preventative maintenance strategies, this strategy could prove to be very costly if left unmanaged.
The unscheduled arrival of breakdowns makes prediction of the expected capacity availability in the system highly uncertain, making planning and scheduling a difficult task. And as a consequence crew planning for the maintenance department also becomes difficult for this reason.
Maintenance forecasting and capacity planning are two important functions for the design of an effective maintenance system. Maintenance forecasting comprises the estimation and prediction of the maintenance load. The maintenance load drives the whole maintenance system.
The ISO 55001 standard address the requirements for effective planning and maintenance support in clause 6 and clause 7 respectively.
ISO 55001 Clause 6: Planning addresses the requirement for identifying all asset related risks with plans on how to treat them. (ISO 55001:2014)
ISO 55001 Clause 7: Support addresses the requirement to consider the resources and competencies needed to establish, improve, maintain and continually improve the asset management system.(ISO 55001:2014)
In conclusion, by effective planning of breakdown maintenance work, an organisation can achieve better control of its maintenance budget and with some degree of certainty be able to reduce lost time incurred due to random breakdown occurrences. For organisations mostly engaged in breakdown maintenance one could make use of the Queuing Theory to better manage the effects of machine breakdowns.
The Queuing Theory was developed to provide models to predict behaviour of systems that attempt to provide service for randomly arising events. The essence of queuing theory is that it takes into account the randomness of the arrival process in this case the breakdown event and the randomness of the maintenance service process. This approach would be most ideal for effectively planning breakdown maintenance activities.
Machine breakdowns or maintenance services patterns can follow one or more of the Theoretical defined Patterns. It is often assumed that machine breakdown patterns follow the Poisson's distribution and maintenance servicers follows Negative Exponential pattern, it could also be assumed that they follow Null-Hypothesis or Normal or Binomial Distribution pattern, hence it becomes necessary to study all the parameters involved.
This can be valuable for an organisation, because by knowing all information related to the performance of it machines, it is more effective and easier for them to plan their maintenance resources in the future, more so in a breakdown maintenance environment.
References
- B. Mahadevan, E 2010, Operations Management: Theory and Practice, 2nd Edition, Pearsons
- Salih O. Duffuaa, A. Raouf, John Dixon Campbell, 1999, Planning and Control of Maintenance Systems: Modeling and Analysis, John Wiley and Sons
- ISO 55001:2014, Asset management — Management systems — Requirements
- ISO 55002:2014, Asset management — Management systems — Guidelines on the application of ISO 55001
Sales | Business Advisory | MSc. MBA
5 年Hi Lizile Good one, maybe one could expand or give the article a sequence with some application examples and more in context theory to allow for a broader discussion. Cheers Kenji
Asset Management | Metals, Minerals & Mining | Industry X.0
5 年Great article Lizile Xulu. Thankfully there is something called World Class Reactive Maintenance which I was fortunate to witness during one of my engagements
Business Owner at Gqunsu Engineering Pty Ltd
5 年The last thing i would like to ask is, did you work with real data to actually observe the distribution of the random arrival rates, say number of breakdown failures per day, and observe the shape of the normalised histogram and conjecture a statistical distribution that actually fits the practical data of arrival rates instead of assuming a Poisson distribution? It could be Poisson, but sometimes it is best to prove this first using Hypothesis testing and then continue from there...But nonetheless i like this...this is a good job bafo...ngiyawuncoma umsebenzi owenzile la bafo!
Business Owner at Gqunsu Engineering Pty Ltd
5 年Yes...i already knew based on the queueing theory application that you were most likely trying to estimate the delay time to sort out the breakdowns as you can estimate the number of servicers you need for a targeted average service delay time. So your maintenance breakdowns among your equipment is also assumed independent...because some equipment failures may cause other equipment to fail too...So this study restricts the assessment to independent equipment breakdowns which are totally de-correlated?
Asset Performance Management | Certified Six Sigma Coach | Certified Reliability ABC Practitioner
5 年In this application the service rates would be mutually exclusive so I would not have more that one servicer on a breakdown. Merely for the reason that the objective is of the study is get average service time in order to estimate the amount maintenance personnel required at any given day.