by R.T. "Chris" Christensen

One of the most useful tools we have at our fingertips for managing maintenance inventory is the ability to forecast the future. Unfortunately, forecasting is also one of the most unreliable tools we have. And because of the unreliability, we need to be extra careful. In forecasting, we try to predict the future so we can minimize the risk of having an unexpected event occur that we might not like. In the current issue of MRO Today, I discussed the forecasting tools and what they can do for you.

Now on the Web, I have some thoughts for you to consider when you look to forecast inventory requirements and face what we call “non-forecastable items.” Under such a scenario, you may wonder if the forecast tools will work after all.

I can’t tell you the number of times I've had people ask me, "What's the best forecast software to purchase?" In one case, a maintenance manager was looking for the software to do this forecast for him. He wanted software to be able to predict when that failure would occur and shut the machine down at the last possible instant before failure. I told him the software he was looking for didn’t exist. Obviously, he did not like that answer. He had told his corporation's higher-ups that he could increase the utilization of those furnaces by managing the outages better and had the green light to get the software he needed to get the job done. All I asked him to do was tell me the hours in service and the date and day that the furnace would fail. With that information, I could give him the software package that would work for him. He thought I was nuts.

He didn’t realize that he was looking right into the weak spot of the software. All that any software package does is take the mean time to failure, look at the parts needed to do the repair and measure that against the planned hours of equipment operation. Then the software does the math to calculate the day and date the next planned outage might occur and base that on historical averages. I asked him to tell me the exact time to failure to load into the forecast equation. He couldn’t give me that information because he didn’t know it. He also could not predict to any degree of accuracy when failure would occur. He knew “that the software could do that calculation for him.” He didn’t understand that he was the one that had to get the information and enter it into the software database so the computer could do the math. Therefore, he thought I was nuts because I didn’t know of any systems that already had the service level database already installed.

I simply wished this gentleman good luck in his search.  I am sure he has had nothing but bad things to say about me ever since. So be it.

The point here is that if you are looking at a forecast tool, realize that it only works on historical data and uses your input to forecast future events. And, it gives you answers in averages. Averages never happen. An average is nothing more than the earliest and the latest an event has happened. Therefore, averages never happen in real life. The average just gives you a clue of when you might expect something to happen. If you have no historical data and/or you cannot provide information on failure and use rates, the software won’t work for you.

R.T. "Chris" Christensen is the director of the University of Wisconsin School of Business' operations management program. If you have an inventory management question, contact Coach Christensen by phone at 608-441-7326 or e-mail cchristensen@execed.bus.wisc.edu.

Back to top

Back to MRO Coach archives