Article 3. How to Determine the Correct Staffing Level for a Function
Industrial engineers are trained for developing efficient manufacturing processes especially balancing assembly lines and manpower loading. They seldom venture into determining the personnel requirements for the rest of the company. As a result most companies are over staffed in virtually all areas except in manufacturing processes. The following example is of a company wide implementation made by a Chicago management consulting firm which I was affiliated with. The client company was a well-known forklift manufacturer located in the northern mid-west. The company was under competition from foreign competitors and had to resort to extreme measures to survive. My assignment was in the Production Control Department and was to analyze the function: “servicing manufacturing“. Several forklift drivers were involved in the transporting of parts on a timely basis to the manufacturing areas either to and from individual milling machines or directly to the assembly line of a particular forklift product.
My approach was to time study all the activities necessary to the servicing of production. I followed and timed each forklift driver into the storage yards and counted the number of parts loaded on each pallet for each trip making sure that I had missed nothing. The number of parts to be loaded on the pallets was specified by the Production Control Planning System. My job was to be sure that the pallets were fully loaded or contained all of the parts required. At the assembly line the pallets of parts were placed near the assembly line so that they could be easily moved to the assembly line worker as needed. This study of the manufacturing areas took about two week
Now lets do the data analysis. I needed to determine what “Product” was being produced by the Function: “Servicing Manufacturing”. This appears to be difficult but what you must do is to break down all of the processes observed and find its smallest element. In other words this element or a multiple of it can be found in all the processes of this function. Since all processes involved a forklift I found that the shortest process was when a forklift moved a pallet from its dropped location near the assembly line to the assembly line. I called this process a forklift move and the Product became “Move”, which I determined to be one minute long. This was very convenient because the entire analysis was done in minutes. If the Product “Move” had been five minutes long you would simply divide the total time for each process by five to get the number of moves. Since each pallet moved had parts on it for more than one forklift and since I had counted and recorded the number of parts on all of the pallets we can analyze each pallet to determine if more than one part is required for each forklift. Adding the total time for the forklift driver to go to the storage area count the number of parts needed put them on a particular pallet and bring the pallet to the assembly area gives us the time to the drop-off place near the assembly line. To that we add the one minute required for a second forklift driver to bring the pallet to the assembly line worker. We divide this total time by the number of parts moved (adjusted for the number of parts required per forklift). This process is repeated for all the pallets. When added together we know the total time spent in servicing this particular forklift product.
We cannot assume that all work will be performed at the standard 100% rate so we assume the actual rate is closer to 75% which adds 25% more time to the total. Multiplying this total adjusted time to service one forklift product by the current production rate (forklifts per month) gives us the adjusted time spent by all the forklift drivers for the month. When we convert the time in minutes to hours and divide this by the standard hours worked per month by a forklift driver we arrive at the total number of forklift drivers required at this particular production rate.
Given the total adjusted time to service one forklift product we can calculate
Staffing levels for any manufacturing rate. It was also determined that since all the forklift products contained nearly the same number of parts the analysis could be used for all of the companies forklift products made on this particular production line.
Now lets review the actual results of this analysis. During the study I staid focused on the processes being performed and the time actually being spent. But since the analysis was being done in the cold of winter I noticed that every time I stopped by one of the four or five warming huts five or six people would get up and leave. Since I had no idea what these people were supposed to be doing I paid little attention to them. The results were shocking to the company senior management when they discovered that the actual number of employees required could be reduced by sixty with an annual savings adjusted for today’s dollars of more than one million dollars. If I had not conducted the study myself I probably would not have believed the results.
