Reliability Engineering – a key to engineering excellence (original) (raw)

Reliability Engineering - a key to engineering excellence

BW Botha
Department of Mechanical and Aeronautical Engineering
School of Engineering, Univ. of Pretoria, Private Bag X20, Hatfield, South Africa, 0028
e-mail: bwbotha@up.ac.za

Abstract

A change in focus has over the past decade seen South African maintenance and reliability expertise disappear to a large extent. This is inter alia due to a misunderstanding of the value and purpose of reliability engineering. It is often seen as an unwanted expense rather than the essential contributor to engineering excellence that it is. It is not as visible as an aesthetically or technically appealing design. “Designer arrogance” often further obstructs logic thinking associated with reliability engineering resulting in an appealing, but less than optimum design. What is often overlooked is that, although reliability engineering does seemingly entail “unwanted cost”, the cost is essential and significantly lower than the cost of non-implementation. Pressures for higher availability have fuelled an increased focus towards maintenance in an effort to improve the availability of systems. However, it is not always realized that traditional maintenance is a reactive and cost intensive way to try and ensure this availability. As reliability engineering entails the identification and planning of activities to ensure that equipment will perform in a predictable manner when you need it for as long as you need it, it includes more than maintenance and forms one of the two key elements of availability, the other being maintainability, and not maintenance per se.

A shift in focus since the 90 's resulted in maintenance and reliability often being neglected in an attempt to reduce short term expenses. Although effective in the short term, the cost to repair or replace poorly maintained equipment increases significantly with huge impacts on resources, availability and subsequently net profit. The shift in focus, together with knowledge of the typical longer term outcome that can be expected, resulted in many of the maintenance and reliability expertise leaving the larger companies either as consultants or to pursue other alternatives, both locally and internationally. Replacing this expertise is, however, no trivial process as the training of juniors in preparation for these positions were not sustained often due to blatant ignorance and/or the perceived cost at the stage. Pressure is further increased due to the shortage of resources to keep production running, often resulting in “fire fighting” removing the “luxury” of making resources available for retraining. This resulted in a significant loss in the engineering capability to ensure equipment reliability and maintainability, thus negatively impacting availability. As with most neglected maintenance activities the impact is only realized once it is too late. This raises the question why the interest in one of the key elements to engineering success has dwindled to these levels despite numerous publications on these topics. The global need is evident by the number of universities and institutions having developed strong research programs in reliability and maintenance over the years (Hines 2001). Even first world countries such as America still have as much as 55%55 \% reactive maintenance, 33%33 \% preventive and 13%13 \% predictive (Stephens 2010). The question, however, is what should be done to change it.

Defining reliability engineering

A major contributing factor is to the reduced interest is the misunderstanding of the aim of reliability engineering. Due to its “supportive” nature it is often seen as unimportant when measured against the more tangible design engineering. The first step towards successful implementation of an effective reliability program is defining what reliability engineering is. Reliability defined as “the probability that an item/system/person will perform a specified function under specified conditions for a predefined period of time or number of cycles” (O’Connor, 2002). Reliability engineering, however, can be defined as all the activities to ensure that equipment will exhibit the required or stated reliability. It therefore includes far more than simply performing maintenance. Reactive maintenance, as commonly applied, addresses the symptoms of things that have already gone wrong. Reliability engineering is about proactively identifying what needs to be done to prevent things from going wrong or, when they cannot be prevented, limit the consequence thereof. Although strongly supported by Reliability Centered Maintenance (or RCM) as advocated by John Moubray (Moubray, 1997) and Nowlan & Heap (Nowlan, 1987), it is much more than only RCM. It forms the basis of a successful integrated approach supporting the current drive towards asset management.

The role of Reliability Engineering in Engineering Excellence

Management buy-in

The impact of reliability engineering activities on the larger picture is not always clearly quantifiable and therefore not as visible as an aesthetically or technically appealing design. The effect of implemented actions is often only visible some eighteen months down the line. This often leads to the implementation of less effective actions with smaller, but shorter term benefits. The real benefit of ensuring optimum availability throughout the product or system life is therefore often missed. Without the necessary knowledge, reliability engineering is then often perceived by management as an unwanted expense rather than the essential contributor to engineering excellence that it is. They overlook the fact that the cost is essential and significantly lower than the often hidden cost of non-implementation (Mitchell 2007). Although some support for maintenance has been created by the discomfort of the disruption of essential services, it is not always realized that traditional maintenance is a reactive and cost intensive way to ensure this availability. Addressing the symptoms rather than the cause increases the pressure on resources and reduces the profit potential of the company significantly.

The importance of reliability engineering in planning all reliability related activities suggests interaction between all parties contributing to the success of a project right from project initiation, supporting the concept of an integrated approach. However, such an integrated approach involves different sections of the company therefore requiring management buy-in supported by sound experience of different engineering aspects, experience typically associated with more senior personnel often in middle and senior management. Management also needs to understand that it is not easy to quantify the value of a sound reliability program, but when implemented effectively, the benefits obtained are normally orders of magnitude larger than the cost (Bloom, 2006).

Although the principles being suggested are by no means new, companies still fail to implement them mainly due to ignorance. Companies that made the mind shift in management have ample proof of the success. This includes industry leaders such as Toyota, DaimlerChrysler, Boeing and many more. For this reason the single most important factor for successful implementation of reliability engineering is a sound understanding of the principles and benefits by management (Stephens 2010). Reliability engineering, like quality, therefore needs to become a way of living communicated down from management through visible support and implementation. If this is not achieved, efforts by individuals often end up in wasting valuable time and money.

Once management buy-in has been achieved, there are a number of essential factors that should be addressed in order in order to successfully implement a reliability program.

Define your goal

The reliability program will be defined by what the company’s goal is. If your goal is to maximise profit in order to sell the company in the short term your focus could be on forcing production at the expense of reliability as you will be selling the company off before the equipment needs replacing, making it the new owner’s problem. If your aim is to build a legacy in a company your focus should be on running at the optimum production rate to optimize life cycle cost over the life of the equipment. This requires careful planning of the most suitable reliability engineering activities. It is therefore important for management to understand their importance in defining the company and reliability program goal. Reliability tasks can be either of a management, engineering or accounting nature, but needs to be defined to meet the goal. All relevant parties should clearly understand the goal together with the relevant tasks required to achieve it. Without a clear goal the activities will lack focus. The simpler the goal, the easier to focus activities towards achieving it. The aim should not be to obtain a discrete and tangible ROI, but to develop a culture to reliably support the focus of management.

Time it right

Due to the misunderstanding of the role of reliability engineering and commercial and financial pressures, reliability engineering expertise are often only included in the project team once reliability issues become apparent during initial testing or even operation. Even when they are involved earlier, the pressure often prevents the “luxury” of allowing sufficient time for reliability activities. The pressure from project managers to produce some hardware to impress customers often contributes to the team falling into the trap of concurrent engineering. Although the iterative nature of larger projects does allow some concurrent engineering, early reliability engineering activities are important to identify the “showstoppers” which should be addressed first. These are the activities which, if no viable solution can be found, will result in either not meeting customer requirements or even

termination of the project. The sheer thought of the outcome often contributes to designers and project managers ignoring the “difficult” questions “until we cross that bridge”. However, such questions will keep on re-appearing until they are addressed, but at an increased cost for each phase that it is postponed to. The " 10 x " rule, or even " 100 x " for the more complex systems today, describes the relation between the cost of addressing the problem when identified or in the next phase. The sooner it is answered, the cheaper it is (Barnard June 2008). If the answer suggests termination, it will remain the answer but at a significant additional cost the later it is realized. The best time to make a design change or terminate is therefore “when you can still click on it and delete it”.

The valuable guidance in protecting the company against hidden requirements during the acquisition phase makes it essential to include reliability engineer right from contract negotiations, but definitely no later than the early development phases. This in general, however, is not the case. As an example there might not be anything wrong with the originally intended design, but the designer might be oblivious to the specific need for a different material. The choice of an alternative material could affect the manufacturing process to be used which could in turn require a different design layout to ensure optimum reliability for the specific application. However, if the design has been completed to detail level before realizing this need it is often extremely difficult and costly to change the design. Time and cost pressures then often contribute to the design being manipulated in an attempt to satisfy requirements. This normally impacts on the reliability and even functionality of the modified end product. Modifying or even redoing the design at these late stages then normally results in overrunning of both cost and time budgets. In the increasingly tight financial environment more emphasis needs to be put on getting the design right the first time. Unfortunately many companies still see these financial pressures as a reason to focus more on getting the design right quickly than getting the right design.

Studies have shown than up to 95%95 \% of the total cost of ownership is determined before it is put into use (Ray). Once manufactured and installed there is very little the reliability engineering process can do to improve either reliability or maintainability and it becomes a maintenance function to try and obtain the required availability after commissioning and mostly at increased cost. This supports the need to have the reliability engineer involved from the early stages of the project. In an attempt to counter this neglect companies incorrectly apply RCM during operation to identify the most suitable maintenance activity rather than ensuring the optimization of the design during the development process as it was intended. Commercial pressures also see all kinds of variations of RCM being applied in order to get the equipment back in operation as soon as possible. However, this results in a quick fix rather than addressing the real cause of the failure. It often also merely shifts the problem to the next piece of equipment. Applying a streamlined RCM due to resource pressures is like taking insurance for travelling on the highway only. It leaves you with a false sense of security making it even more dangerous than doing nothing at all.

Focus on the important aspects first

Human reliability

Various studies indicate human error to be by far the most dominating cause of failures. Van Cott already indicated in the 60’s that human factors are the most significantly contributor to failures. More recent studies have shown this not to have changed much. As a matter of fact, improved technology has reduced equipment failures and increased the impact of human factors. Some of these studies show human factors to contribute to 75−96%75-96 \% of

marine accidents (Rothblum, 2002), 85% of railway accidents (Agarwal, 2005), more than 90%90 \% of all fatal accidents in Australia (Feyer, 2011), 62%62 \% of accidents in the highly regulated nuclear industry (NUREG, 2001), 70-80% of aviation accidents (Shappel & Wiegman, 2004), half of the outage time in the telephone industry (Brown, 2001), etc.

Despite these figures, most of the effort over the years has gone into improving the reliability of systems and products to extremely high levels. This is due to the relative ease of improving and predicting component reliability compared to human reliability. The main focus is towards finding ways to eliminate or limit human interaction with systems requiring high reliability. This is evident in overriding safety systems in most of our daily life applications. However, human reliability offers the largest opportunity for improving overall system reliability. Increasing cost and diminishing returns to improve component reliability has resulted in an increased focus towards increasing human reliability. This includes creating awareness with operators and management at all levels as to the impact of human reliability on overall efficiency. It does not help having a plant with an extremely high reliability being rendered completely unproductive due to the operator not showing up at work. This means that alternative measures need to be considered to either ensure that the operator does show up or that an alternative operator is trained, once again decisions to be made by management. The impact on system reliability has seen an increased focus in developing models to predict human reliability. However, this is still by no means being addressed to the same level as the impact it has. This is supported by the figures by Tait suggesting a 26%26 \% improvement in human and cultural influence to be the largest single contributor to improving reliability and maintenance over the next ten years (Tait 2010).

Identification of the need to model human reliability more accurately has resulted in various mathematical models being developed (Chandler, 2006). However, companies are faced with a number of differences in workforce compilation based on the level of technical ability en technology used. This will necessarily impact on the actual format of the mathematical models applicable to the different industries. Although the models can serve as a good basis, they cannot simply be adopted for the South African industry and need to be reviewed and revised appropriately. However, the principles followed in deriving them could still be largely applicable.

Focus on the asset, not the process

Shortcomings in traditional maintenance methods have resulted in the development of various predictive maintenance techniques, each one exploiting shortcomings of the other. Although these methods are mostly effective when applied correctly, they are often used completely inappropriately, either due to a lack of knowledge or suppliers trying to bias potential clients towards their technology. This not only wastes money, but can result in extremely dangerous situations due to a false sense of security. It is concerning to see the number of companies still supporting a specific technology rather than identifying the most suitable methodology (Davis 2010).

The trend of effective companies clearly indicates a shift towards a risk based approach. Initiated by Nolan and Heap in the 60’s already, RCM, intended as a process during the development phase, identifies the potential asset failure mode first after which a sound reliability strategy is applied with respect to the most suitable technology and frequency to apply as well as the impact on overall life cycle cost. This approach does take more time initially, but the benefit quickly covers the expense (Bloom 2006). A variation on the classic approach is to prioritize assets according to the consequence of the failure and a risk priority number to identify which assets to address first. The trend is to focus on the 20%20 \% assets

indicating the highest risk. A common error, however, is that once these risks have been addressed no further work is done on the remaining assets and resources are utilized for other activities. It is important to note that the decision not to address assets in detail should not be determined by its position relative to other assets, but whether the calculated risk exceeds a predefined risk level or not.

Key elements in Reliability Engineering

To measure is to know

Measure the right things

The well-known quote “To measure is to know” of Lord Kelvin is often exploited by companies marketing their capabilities or equipment to their less informed customers. Although this is true, measuring the wrong things costs you money without telling you what you need to know. Over time various parameters have been measured in an attempt to quantify performance. The result is, however, not always the desired one. In some cases operators feel exposed resulting in a negative effect in the long term. In others the process of monitoring is very time-consuming resulting in a loss of production or resistance to implement. People respond to the way they are measured. If you measure output, then the operator will tend to push output at the cost of another parameter e.g. quality. If you measure maintenance expense the operator will drive maintenance costs down at the expense of reliability. It is therefore extremely important that the right parameters are identified and measured correctly. Although there are generic parameters, the set of parameters for measuring overall performance will be process and company specific. If your aim is output you will measure numbers whereas if your aim is less comebacks you will measure quality. It may also be location and season specific. What exactly you measure is therefore far less important than the process used to decide what to measure. Spending time in ensuring that the right parameters are measured will therefore be money well spent in the long run. However, the team needs to understand and support this notion. Although care should be taken that the process is not exhausted, the importance of good planning should never be underestimated.

A common mistake made by engineering management is focusing on measuring the performance of the individual. This results in the employee focusing on the activities that

offer the maximum potential for personal gain and hiding any information that could reflect badly, even if this is to the detriment of the process or company. Supporting this culture often sees the employee driving short term goals and leaving as soon as a better opportunity presents itself, thus impacting on continuity and overall performance figures. This in return sees management requesting specific skills to ensure production when appointing personnel rather than skills for relevant problem solving. Measuring the team shifts the focus away from the individual to focus on tasks to the benefit on the team and supporting activities that increases the overall team performance. Problem areas are far more regularly exposed for early rectification, thus supporting a “team” culture. This does not mean that individual appraisal is unimportant, but that attention to balance should be included.

Targets

Reliability analysis is often performed and the results published in a report, but very seldom include target levels. Without these targets there is no way to determine whether the measured values are good or bad values. This could be because the company has not defined targets, but it is often because it is incorrectly assumed that everyone knows the targets. This often results in the achieved values being used to define the target leaving no incentive to improve (Dunn 2012). The target should therefore be a realistically achievable target to aim for and benchmarked to industry standards. It should be visible and reported on together with achieved values, should include both the value as well as a tolerance of acceptable deviations and should be accompanied by suitable actions to be implemented when not met. Achieving this goal should never create complacency. Once the target is met consistently it should be reviewed and increased realistically in an effort to improve overall performance. A target is therefore essential to ensure control and serve as incentive for improvement.

Analyse

A trend appearing in industry is that an ever increasing amount of information is being gathered, but with very little analysis being done. This is partly due to the availability of improved monitoring technology and the lack of capable analysts. Huge amounts are currently being spent on monitoring equipment to gather information while far less attention is paid to understanding what the suitable action would be to improve the situation. It is simple, if you do not have a reason to measure, do not do it. If you do, then use a suitable technique together with sound analysis to determine what the measurements are actually indicating. Simple measuring techniques with sound analysis capabilities can go a lot further in improving reliability than state of the art equipment without analysis. Measuring without doing anything is a waste of money. However, analysis results can be misleading and dangerous if the analysis is not performed correctly. It is therefore extremely important to ensure that personnel responsible for analyses have competent and capable resources.

Key Performance Indicators

A key performance indicator (KPI) is a parameter putting measured performance values in context. It indicates whether the actions taken to improve overall performance have been successful. It therefore needs to include targets as well as allowable levels of deviation. The focus of KPIs should be to ensure that a company is reducing cost and risk while improving performance. Evolution of the focus on asset management has seen an increased emphasis on performance related metrics and KPI’s. This is confirmed by Tait expecting performance to be the most important set of maintenance metrics to be addressed in the next 10 yrs by a significant margin ( 30%30 \% ). Two very important indicators are the overall equipment efficiency (indicating efficiency of equipment utilization) and the overall labour efficiency (indicating

Overall Equipment Efficiency

Overall equipment efficiency (OEE) has become the accepted standard for measuring equipment performance and utilization. It is defined as the product of availability, performance and quality (Vorne) and reduces complex problems to simple presentation of information assisting in easy and accurate decision making.

Overall Labour Efficiency

The realization of the impact of human reliability lately has seen the development of Overall Labour Effectiveness measuring availability, performance and quality of the workforce (Kronos). However, it is mostly based on the assumption that the operator is asked to perform standardized tasks and is utilized the whole time, i.e. idle time waiting for the next components, the machine to be repaired or scheduled breaks are excluded. Variations can be used to include these times and re-assign employees to other tasks. Performance indicates whether the employee performs the work within standard or allocated times and is calculated by the actual output over the expected output. Strong contributors to improved performance are the level of skill and effective training.

OLE is a valuable tool for measuring productivity and identifying areas to improve, be it operator efficiency or process related. Defining variations of the above can offer a good indication as to the effective utilization of employees as well as opportunity for re-allocation. More importantly, however, it can be used to obtain valuable information for early identification of subtle issues affecting personnel attitude and performance.

MTBF & MTTR

Two other important performance metrics are the MTBF and MTTR. MTBF or mean time between failure is an indication as to the reliability of the equipment, i.e. how long the equipment will operate without failure while MTTR or mean time to repair is an indication as to the maintainability, i.e. how long it takes to return the item to an operational state. As the components of availability, measuring the MTBF and MTTR gives an indication as to the scope for improving the achieved availability. However, both reliability and maintainability are related to the design of the equipment and therefore need to be addressed during development. Once the design is fixed, these values are, for all intent and purpose, fixed. This supports the drive towards DfX, or then design for reliability, maintainability, availability and sustainability, thus focusing on addressing the various issues during the development of the product.

Do NOT play the numbers game

Reliability engineering is often associated with calculating a reliability figure or measuring performance for determining bonuses. However, reliability engineering is far more than only calculating figures, in fact most reliability problems are solved with very limited detailed calculation. Reliability experts often predict reliability values pretty accurately and the more effort is put into the models, the more accurate it becomes. However, these models seldom consider the effect of variations in the system or process resulting in the figure calculated often deviating significantly from the actual performance measured. Figures should therefore merely be used as an indication as to what could be expected if all other conditions were exactly as simulated. It does not mean that the value has no value, it simply means it should be used wisely. It should not be seen as a value to measure against, but as a value to compare

options and serve as guidance during the decision making process. Sensitivity of the system performance to the interaction and variation of parameters are far more important than the actual value itself.

Barnard suggests that we should distinguish between reliability engineering and reliability accounting with the latter lacking the fundamental aspect of paying any attention to the detail (Barnard 2009). Numbers are often used to create the impression of an effective reliability program, but as with quality management, the purpose should not be to impress, but to identify areas of concern and improvement to ensure a positive impact on the life cycle cost and overall company efficiency. If the numbers become more important the company has an incorrect focus indicating an “immature” understanding of reliability engineering. Without a solid understanding, reliability “expertise” uses the reliability numbers to justify their existence rather than positively impacting the overall performance.

Use and educate the right people

Maintenance problems are often addressed by increasing personnel numbers to counter for “fire fighting” (Turner 2001). However, pressures on resources, especially economic resources, limit this “luxury”. It therefore is becoming more important to reduce maintenance requirements through proper planning based on reliability engineering and sound “design for” principles. If done correctly, it shifts the focus to proactive maintenance reducing the requirement for maintenance personnel. This shift focuses the attention on the life cycle cost often supporting less “more expensive” reliability engineers to ensure that the right things are done rather than more, lower cost “fire fighters”.

Reliability is the responsibility of the design team. However, as the designer does not have the time to perform formal reliability activities together with the normal design activities and he/she cannot review his/her own design work effectively, the responsibility is often “outsourced” to a reliability engineer. Inexperienced design engineers often feel threatened by the fact that their “incompetence” can be exposed by others. They therefore often tend to “hide their incompetence” through “arrogance” and excluding the expertise that could make valuable contributions to improve the design. This is partly due to their education process measuring the individual on focused aspects of specialized fields, thus creating the impression that overlooked aspects is seen as incompetence rather than lack of experience. It is therefore important that young engineers understand the purpose of a design team and that their value is by no means threatened by inputs from others. It should rather be seen as security to ensure a design addresses all potential failure modes and exceeding customer expectations. The importance of team work, design reviews and experience of others should therefore never be underestimated.

In financial tighter times, companies often tend to allocate junior engineers as reliability engineers in one of the Maintenance, Quality or Logistics departments to do the activities “not important enough to keep senior personnel engaged with”. As they gain experience they are moved out as the pressure on resources requires experienced personnel in other departments. This not only creates the impression that reliability engineering is less important than other engineering activities, but seriously impacts on any continuity of efforts. This is the best indication that the company does not understand the purpose and importance of reliability engineering. Applying reliability engineering correctly requires sound engineering knowledge of the equipment to be designed and therefore requires guidance from mature engineering expertise. The reliability engineer is responsible for the continuity in requirements and identification of potential failures on various aspects addressed by multiple designers. It is therefore important that the reliability engineer forms part of the design team and is not

allocated to service departments as is normally done. It is also important that the team understands that reliability, and therefore the identification of any potential problem, is the responsibility of the team and not only that of the reliability engineer. It is a continuous effort throughout the development of the product. The reliability engineer is merely the responsible person for capturing the relevant information and ensuring that reliability engineering activities are addressed and coordinated when required.

Although someone needs to take charge, a reliability effort revolving around one person’s passion is doomed to failure once that person decides to leave the company. It is therefore extremely important to ensure that sufficient knowledge transfer occurs between employees to ensure a wide reliability base and continuity should key personnel no longer be available.

It is therefore important that management and project managers understand the value that a good reliability engineer can add to their team.

Adopt the right purchasing approach

Purchasing in larger companies is typically the responsibility of the purchasing department. However, the purchasers are often employed based on their ability to source the cheapest supplier for the required item i.e. reducing capital expenditure. A lack of technical knowledge often means that the purchaser does not understand the impact of small variations in specification on the performance and overall life cycle cost. A cheaper item may result in a significant reduction in item cost, but the associated loss of income due to an additional shutdown during the operation of the plant may exceed the additional capital cost by orders of magnitude. It is therefore essential that the purchasing personnel understands the importance of preferred suppliers and the role of the designer and systems engineer in specifying suitable equipment and suppliers and work together with them to obtain the required overall performance at the lowest possible cost.

The latest trend in industry is to purchase items on a service contract where the supplier is responsible for maintaining the product for a predefined period of time with relevant penalty clauses being invoked in case of unplanned shutdowns or reduced performance. This means that the customers are focusing on paying for results rather than promises. This has seen suppliers giving increased attention to reduce maintenance through increased application of Design for Reliability principles during the development phase, something neglected or ignored previously due to the negative impact on competitive product pricing. However, shifting responsibility of maintenance to the supplier now makes it cheaper for the supplier to address it during development rather than during operation.

Apply Reliability Engineering for the right reason

Reliability engineering is all about the scientific application of engineering decisions to improve overall performance. An effective reliability program will be integrated into management strategy to support a “way of living”. Reliability engineering should therefore result in a companywide mind shift towards the development of an iterative way of thinking and not solving isolated incidents. It should not be done to impress customers, but to improve the company’s performance. For this it should be integrated into every decision process and not performed as an add-on. The aim is to move from a reactive “fire-fighting” mode to a proactive mode. Identification of maintenance activities are therefore not to reduce maintenance, but to identify the right maintenance to do to ensure optimum life cycle cost. Failure modes should not be fixed with the aim of surviving, but should be addressed to improve the process or product design. The customer will benefit and return because they will experience the efficiency in all aspects of service.

In order to achieve this mind shift it is important to educate employees at all levels of the company as to the importance of reliability engineering. This should be supported by a clearly visible message from senior management as to their support. Focusing on improving the process or product will shift the focus to the development phase as the only time to effectively address reliability and maintainability. This will require the involvement of senior expertise early on in the development process.

Reliability engineering activities and targets should not be implemented as a policing action to threaten low employee performance as this will result in employees tending to cover up important information w.r.t. failures or deficiencies, i.e. defeating the effort. Attention should rather be focused on creating a supportive environment for improving performance and giving recognition to employees identifying potential opportunities for improvement. The value of such recognition should never be underestimated.

In conclusion

The aim of reliability engineering is therefore not to impress anyone, nor to blame anyone. Reliability engineering’s primary goal is to identify potential problem areas as early as possible in order to prevent unplanned, costly failures and ensure that equipment will perform in a predictable manner when you need it for as long as you need it. It cannot be used to improve reliability of a poor design once in operation. It should be implemented early in the project, but no later than the early development phases in order to address all aspects of the product development ensuring optimum reliability and maintainability for the life of the plant. The aim should be “to do the right things right the first time”. It requires a corporate mind shift and therefore management buy-in is essential for successful implementation of a reliability engineering program. As companies realize the value and shift to more regularly implementing reliability engineering, companies that want to ensure their continued competitive place in the market will have no choice but to follow suit. However, this will require proactive educational programmes similar to the initiative between industry and the University of Pretoria to ensure suitable training well in advance of the actual need.

References

Agarwal, A, “Rail Accidents due to Human Errors - Indian Railways Experience”, International Railway Safety Conference, 2005, Cape Town, South Africa.

Barnard, RWA, “What is wrong with Reliability Engineering”, 18th 18^{\text {th }} Annual International Symposium of INCOSE, 6th 6^{\text {th }} Biennial European Systems Engineering Conference, Netherlands, June 2008.

Barnard, RWA, “Integration of reliability engineering into product development”, 2nd 2^{\text {nd }} SAIAS Symposium, Stellenbosch, South Africa, September 2008.

Barnard, RWA, “Modern Reliability Engineering Based on Quality Principles”, 2009 ARS, Europe: Barcelona, Spain, 2009.

Bloom, N, “Reliability Centered Maintenance - Implementation made simple”, ISBN 0-07-146069-1, McGraw-Hill, 2006.

Brown, A, Patterson, D, “To Err is Human”, First Easy Workshop, Univ of California, Berkeley, July 2001.

Chandler, FT, Chang, YH, Mosleh, A, Marble, JL, Boring, RL, Gertman, DI, “Human Reliability Analysis Methods, Selection Guidance for NASA”, NASA/OSMA Technical Report, July 2006.

Davis, J, “Predict to Prevent: 7 smart maintenance strategies”, http://www.plantengineering.com , (accessed 03-12-2010).

Dunn, S, “Using performance measures to drive maintenance improvement”, http://www.plantmaintenance.com , (accessed 05-01-2012).

Feyer, A, Williamson, A, “Human Factors in Accident Modelling”, Encyclopedia of Occupational Health and Safety, International Labor Organization, Geneva, 2011.

Hines, J.W., “Maintenance and Reliability Education”, Emerson Process Management Reliability Conference, Nashville, TN, October 22-25, 2001.

Kronos for Manufacturing, “Overall Labor Effectiveness (OLE)”, http://www.workforceinstitute.org/ wp-content/uploads/2008/01/ole-achieving-highly-effective-workforce.pdf, (accessed 14-02-2012).

Lifetime Reliability Solutions, “A common misunderstanding about Reliability Centered Maintenance”, http://www.lifetime-reliability.com/free-articles/maintenance-management/ A_Common_Misunderstanding_about_Reliability_Centred_Maintenance.pdf, (accessed 04-032012).

Mitchel, JS, “Physical Asset Management Handbook”, ISBN 0-9717945-4-5, Clarion Technical Publishers, 2007.

Moubray, J, “Reliability Centered Maintenance”, 2nd 2^{\text {nd }} edition, ISBN 0-978-0831131463, Industrial Press Inc, 1997.

Nowlan, F.S, Heap, H.F., Reliability Centered Maintenance, Report A066-579, 1978
NUREG, “Review of Findings for Human Error Contribution to Risk in Operating Events”, INEEL/EXT-01-01166, US Nuclear Regulatory Commission, August 2001.

O’Connor, P, “Practical Reliability Engineering”, ISBN 978-0-470-97982-2, John Wiley & Sons, 2012.

Ray, D, “What’s the role of the reliability engineer?”, Life Cycle Engineering, Reliableplant, http://www.reliableplant.com/Read/23083/role-reliability-engineer-operations, (accessed 08-082012)

Rothblum, AM, Wheal, D, Withington, S, Shappell, SA, Wiegmann, DA, Boehm, W, and Chaderjian, M., “Human Factors in Incident Investigation and Analysis”, 2nd International Workshop on Human Factors in Offshore Operations (HFW2002), Houston, TX, April, 2002.

Shappel, S, Wiegmann, D, “HFACS Analysis of military and civilian aviation accidents: a North American Comparison”, ISASI, Gold Coast, Australia, 2004.

Stephens, MP, “Productivity and Reliability-Based Maintenance Management”, ISBN 9-781557535924, Purdue University Press, 2010

Tait, A, “Physical Asset Management Benchmarking”, Physical Asset Management Thought Leadership Conference - Mastering the Game, 11 & 12 May 2010, Cape Town, 2010

Turner, S, “Maintenance Analysis of the Future”, International Conference of Maintenance Societies, Melbourne, 2001.

Vorne Learning Centre, “OEE (Overall Equipment Effectiveness)”, http://www.vorne.com , (accessed 14-02-2012).

Biography

Dr Botha started his engineering career as systems engineer at Denel in 1992 and registered as professional engineer with ECSA in 1994. He joined Potchefstroom University in 1994 where he lectured at both under and post graduate level till 2008. In parallel he acted as consultant to PBMR and quality manager to M-Tech Industrial. He completed his Ph.D. in Engineering in 2003 at the North West University. In 2008 he joined PBMR full time as Senior Systems Engineer. After the closing of PBMR in 2010 he joined the University of Pretoria lecturing in Reliability Engineering, Reliability Based Maintenance and Mechanical Design as part of the Maintenance Engineering programme. These courses now form part of the offering of the Centre of Excellence in Engineering Asset Management and Condition Assessment. Further information can be obtained from Dr Botha at +27 -(0)12-420-4570 or bwbotha@up.ac.za or from the website of the Department of Mechanical and Aeronautical Engineering at the University of Pretoria.