Friday, May 18, 2012

Transitional Root Cause Analysis

When I discuss RCA I use a method called Transitional Root Cause Analysis or TRCA for short.
It is made up of 10 tools that can be explained and understood in a very short period of time.
In the next few minutes I will demonstrate both the simplicity and rules for use for 3 of the 10 and explain why we consider them transitional in nature.
In this blog we will use what I categorize as the tree methods. These are three tools that build out into a tree root like structure. So let's take a look: the first one is called the "5 why"  method. It is very common in industry today and is a very basic root cause tool. It is created simply by asking the question why multiple times to create one causal chain. It creates a simple main tap root to build off of.
Now, if we take the 5 why diagram and branch it out by adding more elements at each level then we get a better representation of all of the causes that come together to create an effect. This transition of the 5 why is known as a fault tree. This method allows us to easily see all factors that led to a failure, but sometimes we need to show a bit more information to make the graphic more meaningful. 
If for instance an effect can only occur when all of its causes exist at the same place and in the same moment in time then we use the word "and" at that junction of the roots. If we eliminate either one of the cause then we can eliminate the effect. On the flip side if either of the causes could precipitate the effect then the word "or" would be placed at the junction. This would be read as this or that could cause the effect above. This allows you to see that both possibilities must be addressed to prevent the cause.
These three tree methods transition from one to the next by adding one simple new feature as needed during the root cause process. First, we take a "5 why" and branch it to get the fault tree then we add in the "and and" or "or" to get logic tree. Three powerful tools that build on each other to get you to the lowest cost solution that mitigates the risk.The other Transitional tools work very much in the same way and allow us to use the right tool for the job instead of trying to use a screwdriver as a hammer.
If you would like to learn more about the Transitional method to RCA please send me an email and I will help you along.

Friday, May 11, 2012

Question from the Field: Maintenance Ready Backlog

So what is maintenance ready backlog, and why do I need 2-4 weeks of it?
Maintenance ready backlog represents the work orders that have progressed through your maintenance business process to the point that they have been approved and are planned for execution. These jobs have all of the parts, tools, and work procedures packaged and completely ready.  When we say you need two to four weeks that means that you should have enough work in this stage so that if your crafts had no break-ins or emergencies to do then you could keep them all fully employed for two to four weeks.
You need this amount for a few reasons:
First you want to be able to use these "ready to work" jobs to build your schedule each week. You need at least two weeks selection for operations and maintenance to choose from to be able to create a fully planned schedule that takes into account available equipment downtime and manpower. Your goal is to keep unplanned jobs and jobs without the parts on hand off of the schedule. By doing this you will lower maintenance cost and minimize schedule breaks. If you get below the two week mark you will start to see that you just do not have enough work orders to adequately utilize all of your crafts or the available scheduled downtime and many site are then tempted to pull in unplanned jobs and try to rush the parts and the planning.
Second, You don't need to exceed four weeks because you do not want to have excess inventory and tools tied up in the ready backlog.
To manage this level there needs to be constant communication between the Maintenance Manager, who controls the amount of craft hour available for work completion, and the Planner/Scheduler. If the Maintenance Manager were to approve extra hours of overtime for a few weeks then that will consume the ready backlog at an accelerated pace. The planner would then have to step up the production of job packages or face the possibility that they could run out of ready backlog and have to work on unplanned work.
If you can maintain the level in the band you will find that it allows for a much more effective and effective overall maintenance process. This equals lower maintenance cost for the same or higher levels of reliability.

Thursday, May 3, 2012

Questions from the Field: Wrench Time Studies

Questions: You had mentioned that we should only conduct a wrench time study on planned work.  What is the reasoning behind this?  If they get pulled off of their scheduled job to work on a“sponsored” job, should we terminate the study for that day?  It seems that we have a lot of sponsored work this week and it is somewhat difficult to stay focused on planed/scheduled work.

Answer: That is correct, planned jobs only.

The purpose of the wrench time study is to see how good the planning system is.  If they are pulled off onto sponsored work, the wrench time will likely drop significantly because troubleshooting, looking for parts, looking for prints, and waiting on production are all part of unplanned work but not a part of wrench time.

Of course this is not to say that you cannot continue the study, but it will represent something very different than if they just stayed on planned work.

So it could be as simple as this:

If they stayed on planned work then the wrench time study is a measure of the quality of the planning process.

If they do not stay on planned work, then the wrench time study is a measure the re-activeness of the organization and how well the crafts are able to accomplish unplanned work.

Both of these are valuable, but they represent two totally different things.

Hope this helps.

Andy Page
Allied Reliability Group