In short, ‘root cause analysis’ is intended to prevent the recurrence of defects. Bernard Piovesan, an engineer who lectures in RCA in Australia and New Zealand, says managing the effects of defects is considered in the RCA process to help set priorities for investigation and ROI estimates. “Ultimately it is about stopping things happening again.”
Piovesan adds that RCA is used to varying degrees in industry – usually based on the perceived costs of a problem recurring (i.e. safety, lost production, repair costs etc).
“The need for it to be used is often identified, but the execution of the RCA process is often not done well. The reasons for this are many. Staff can become fixated with symptoms, and dealing with them. Recurrent problems can be assumed to be caused by the same things that may have caused them last time. The urgency of returning to production can cause evidence to be destroyed or not properly examined. 
“Ultimately, to do an RCA properly requires disciplined adherence to the RCA process, objectivity, and ownership of the investigation by a tenacious team leader. In many respects, a successful RCA is like any successfully managed project: the principles and skills needed to achieve a worthwhile conclusion are the same,” says Piovesan.
Barry Kleine, reliability manager, US, New Zealand and Australia for ABB Limited, agrees that RCA execution is often lacking. “I have seen sites where I believe 95 percent of the RCAs done will not return the benefits expected. An issue we frequently see is engineers and managers who believe they ‘know’ what RCA is and how to do it, so won’t invest the time to see what is new regarding the process. There has been a great deal of development in this area recently, and RCAs can be done a lot faster than previously.”
So how should a company go about applying RCA within its operations? Kleine suggests the following steps:
• Get management agreement that the RCA process needs to improve. If some of management thinks the current process is fine, no amount of training will make a difference. Part of this involves ‘sales’ of the issues on site and why a better RCA process will assist business objectives.
• Set up a structure to support and recognise RCA work. Who’s in charge of it on site? How do you know when an RCA is of acceptable quality? How are you tracking the actions implemented, and that the solution does what it should? Even a basic RCA method will be valuable if managed correctly, and a fantastic method will fail if it is not.
• Properly understand the losses on site and show the business case why a topic should be worked on. Most RCA training companies I have seen don’t cover this, or cover it well, as part of the training.
• Give the required training and support to execute a good RCA.
• Sustain momentum by publishing improvements gained.
A common reason why RCA investigations are not completed satisfactorily, suggests Piovesan, “is the habit of organisations to lurch from one crisis to another. 
“Problems that are flavour of the month today, requiring investigation, are replaced by something deemed more important tomorrow – and on it goes. Like a dog chasing its tail, much time is expended for little value in return, because nothing ever gets finished.”

Biggest misunderstandings
“The biggest myth about RCA, says Piovesan, “is that when doing an investigation, if you do A, B and C you’ll find “the” cause that made it all happen!
“I can only recall ONE problem where there was one, single cause. In the majority of cases, there are a combination of factors that are probably contributing to the problem – subject to how much data you have and how well you can believe it. 
“Also, though RCA is a rigorously scripted process, we humans are not robots or computers – we do not necessarily think sequentially and coherently. A good RCA investigation must allow for the creative yet meandering thinking that often produces useful insights and ideas. How the investigation is facilitated is very important for this,” says Piovesan.
The biggest mistake Kleine has seen is not putting the same focus on identifying the problem areas as is spent on doing RCAs. “A fantastic RCA on something that is not a major loss will still only result in low returns. On a site I am currently supporting in root cause analysis, we quantified the losses and agreed the top nine issues. Eight of the nine no-one had looked at before, although ‘RCAs’ were common practice. Everyone was very surprised to see what the true issues were. A lot of time was then saved by stopping RCAs on just anything that came along.”

There’s also a misunderstanding that you should do RCAs to mitigate the same failure occurring again, says Kleine, “which is not quite right. It is about mitigating significant issues that are impacting business goals.” This misunderstanding results in people doing lots of RCAs on all sorts of things believing they need to stop all issues reoccurring, adds Kleine. It also results in people not looking hard enough at the history. “If this thing last failed five years ago, the new one may last five years, so an RCA and mitigating solutions will not show any return for the same period.
“But when asked, no site will accept a five year payback on a project. My success in this topic has been by solving issues which return value in a very short timeframe. Frequency and impact play a significant part in choosing a topic to analyse.”
RCAs are not just for equipment either, says Kleine. It can also be used for people issues, safety issues and business issues. “Anywhere something is happening that you are not happy about, you can use the process.”

A message to management
If senior management are serious about incrementally improving their businesses they must have a strategy for gradually removing defects from their organisation, says Piovesan. “You can’t get better by continuing to do the same things over and over. RCA is a defect elimination tool that’s at the heart of continuous improvement.”
Your people cannot effectively address all issues that occur. Reacting to a 'failure' is often not a good use of your people’s time, says Kleine. “Before starting an RCA, ask yourself: 'Is this the most significant issue I have at this time?' and 'When am I likely to see the benefits from addressing this issue?'. If you are not happy with the replies, repair the equipment and let your people get back to the real issues.
“Pay attention to the results – ask to see the verification,” he says. “I see many good looking RCAs, but the actions are not followed up to see if they are actually done.”
Both Kleine and Piovesan conduct RCA training via the ongoing seminar programme at the Manukau Institute of Technology’s Maintenance and Reliability Centre.