Over the years, maintenance of machines has become hard because of the interdependence of their components. There several papers that have suggested solutions to the existing problems and others provide probable preventive measures for optimal maintenance. Some of the reviewed articles revealed that there are a number of maintenance models with the two mains ones being individual maintenance and group maintenance models. Some of the ways used to search for maintenance modeling articles are internet services or use of published materials unavailable online. Overreliance on internet articles and use of different terminologies make it hard to find the right articles that could be applicable in multiple component maintenances. Maintenance models are categorized depending on the aspects such as economic, stochastic, and structural, and each model has positive and negative impacts in each aspect. On the economic front, multi-component maintenance models are reviewed in two manners that are positive and negative dependence.
Looking at the positive economic dependence, the cost is saved when maintenance is done on joint components than when maintenance is carried out on separate components. Positive economic dependence occurs in two ways; downtime opportunity and economic of scale. Economic of scale is applicable when maintaining components jointly, as it assists in cost reduction. There are several forms of the model of economics of scale such as general economies of scale that cannot change if a single set up cost is introduced. On the other hand, single set up form incorporates both fixed and variable costs with a two-component series system analyzed using this model. Additionally, a multi set up model applies when several set up activities are scheduled to take place at the same time when maintaining several components and the model takes a hierarchical shape. Similarly, downward opportunity is applicable where both corrective and preventive maintenances take place simultaneously.
Negative economic dependence model comes into play when maintaining joint components is more expensive than maintaining them separately. The reason behind the excessive costs in the group maintenance is the problem about balancing between workload and the hired labor, legal and safety barriers and losses in production. K out of n function is used to analyze systems with both negative and positive economic dependences. The system may fall in three subcategories depending on the value of k with respect to the value of n, and these categories are parallel, series, and redundancy. Stochastic dependence occurs when in multi component models that have failure interactions between the components. In this degree of dependence, the condition of one component can affect the condition of another component and the condition can be measured in various parameters.
The components can suffer three types of failures for instance, type I failure, failure type II failure and type III failure. Type I and II failures apply in simplified policies for maintenance where failed components get an immediate replacement. The third failure type considers a reliability system of n components with age and state of the components being used to estimate the reliability of the components. In structural dependence, the failed components must be replaced since the model in this type of dependence is maintenance-dependent and not failure-dependent. Shifting focus to planning horizons of maintenance optimizations there are two categories involved such as finite and infinite horizon planning. In the prevailing trends, several articles are reviewing the applicability of optimal maintenance systems with stochastic dependence. However, some areas of optimal maintenance have low coverage of research work and more evidence is needed in these areas.
A Survey of Maintenance Policies of Deteriorating Systems
Systems used in big industries experience deterioration due to wear and tear, and they are mostly repaired incase of failure. The persistent significance of maintenance has attracted various strategies to improve the reliability of the systems. Models have been classified in various categories to enable individuals identify the system that fits best in optimized maintenance. There are policies that govern the maintenance processes of the system and the two maintenance classes involved are corrective and preventive maintenance. On the other hand, there are a number of maintenance policies that include, age replacement policy whereby a system is replaced after a certain time or after failure. The second policy is periodic PM policy that mainly focuses on the prevention, as the system is maintained at specific time intervals regardless of its failure history and is regularly repaired at intervening periods. The third policy is failure limit policy whereby preventive maintenance is carried out after failure rate is significant and when the rate of reliability of the system deteriorates with a huge margin. In this policy, the system is limited to operating at or over the threshold reliability level.
A table of summary of maintenance models
In sequential PM policy, the maintenance is carried out at uneven intervals of time, as the frequency of maintenance increases with the operation time of the system. Repair limit policy comes into existence when a system fails to operate and cost of repair is estimated. For a repair to take place, the estimated cost must be less than the predetermined limit cost otherwise there would be a complete overhaul of the failed system. Morimura and Makabe developed a repair number counting and reference time policy that allowed the systems to be replaced after certain numbers of failures occurs; otherwise they advocate for minimum repair if the limit number of failures is exceeded. The age-dependent PM policy and periodic PM policy have featured consistently in the single component systems as several articles covering single user systems only feature the two policies.
There are also six policies governing a multi-component system that is stochastically deteriorating in its reliability. Group maintenance policy is the best considered policy in the literatures, as it assists in cutting down operations cost and retains systems reliability for a long time. Furthermore, opportunistic maintenance policy is applicable in multi-component systems where there is interdependence of state of components. Finally, optimal maintenance policies focus to enhance availability of systems and lower the degree of failure and reduce downtime.
In conclusion, optimal maintenance has become important in the recent past, as many articles have increasingly discussing it because it has proved to be economical in several industries. Additionally, several optimal maintenance models govern the type of maintenance that can fit a system in case the repair or replacement is required. The models are categorized with reference to their dependence on certain aspects such as economies and structures of the systems among others. Guiding policies in each category of maintenance are vital, and the policies are fixed to a specific system. Therefore, for the industries to maintain their systems they have to select the best maintenance model and be sure of the policies that go along with the model.
