System maintenance managers should complete a man-hour-data checklist. It will reveal surprising results about accuracy of the data.
Today's automated Army uses a management tool called the standard Army maintenance system (SAMS) to reduce labor costs. To reduce labor costs and maintenance backlogs, SAMS tracks individual and aggregate work orders, repair parts, and man-hour-data trends.
As the Army develops systems to do business better and save money and resources, maintenance managers must have confidence in the accuracy of available man-hour data. In the September-October 1994 issue of Army Logistician, Lieutenant Colonel John R. Hills, Jr., and Lieutenant Colonel (Retired) Michael J. Mannion described integrated sustainment maintenance (ISM), which helps to provide that confidence.
One ISM objective is the factual accounting of work center costs to reflect the true costs of doing business for sustainment maintenance. Because ISM relies on man-hour data uploaded to it from SAMS and other existing systems, the SAMS man-hour accounting process must be accurate for ISM to be successful.
Maintenance officers lack confidence in the accuracy of man-hour accounting data because of errors shop foremen make when they record man-hour data on the work order forms (DA Form 5504) and perceived erros clerks make when they manually transfer man-hour data from the work order forms into SAMS. The reporting of inaccurate man-hour data have several repercussions. First, maintenance managers cannot reduce their backlogs with man-hour schedules adjustments when they lack accurate man-hour data upon which to base decisions. Managers normally shift backlog among shops to expedite repairs. However, if they cannot quantify each shop's backlog by man-hours, managers cannot schedule work efficiently.
Second, inaccurate input, consolidated at the Department of the Army (DA), gives DA a false picture of the man-hours required to repair each type of equipment. These inaccurate man-hour data will cause sustainment maintenance managers at all levels to make inefficient work-load decisions. Also, DA officials may cut or reallocate assigned personnel if they perceive that maintenance companies are mismanaging their allocated man-hours.
Shop foremen accurately record man-hour data on the work order form.
The SAMS system accurately process the data.
All data transfers between SAMS-1 and SAMS-2 occur in the proper sequence.
Having made these assumptions, I started my examination of man-hour data. First, I collected man-hour data found on randomly selected work order forms from the 557th Maintenance Company's completed work order file for 1 January through 24 September 1993. To have enough data to be representative of the man-hours on all the work orders, I needed data from nearly 200 work orders.
To ensure a random selection of data, I asked a stranger to pick a number between 1 and 21. He selected the number 10, so I removed the tenth work order from the closed work order file and proceeded to remove every 21st work order until I had about 200 work orders. From these work orders, I consolidated all the man-hour data on a chart.
Next, I collected SAMS man-hour data corresponding to the random sample of forms I had used and consolidated all these data on a chart. This chart made it easy to enter the data into a computer program in card-column sequence. I used a social science computer software program to help organize and analyze the data. For purposes of this study, I grouped equipment repaired into four types: communications equipment; heater equipment; wheeled automotive equipment; and tracked equipment.
I compared the total man-hours recorded on the forms with the total man-hours in the SAMS system. I then subtracted the total SAMS man-hours from the total man-hours recorded on the work forms. If the deviation was zero, the clerks had input the data accurately. If the deviation was a negative number, the clerks had made an error. A negative value meant SAMS showed more man-hours than the shop foreman recorded on the work form. If the deviation was positive, SAMS showed less man-hours than the shop foreman recorded on the work form. This error would cause DA to reallocate man-hours from the company.
My study showed that, although the difference between SAMS man-hours and work form man-hours ranged from -35 to +9 man-hours, the variation was statistically insignificant. The mean difference in man-hours was -.249 hours. On average, the clerks erroneously added about 15 minutes per job.
Further evidence of the clerks' accuracy in inputting the data was shown by where the frequency of error fell. Approximately 60 percent of work orders examined showed that the clerks made no mistakes inputting man-hours. For another 20 percent of the work orders examined, the deviation between SAMS and the work orders was plus or minus 1 man-hour.
As a former shop officer, I doubted these findings-something must be skewing the figures. Maybe the mistakes the clerks were making could be linked to the type of equipment being repaired.
At this point, I ran a cross-tabulation between the deviation and the type of repair shop for each work order in the bottom and top 10 percent of the frequency range. Of the work orders in the bottom 10 percent, 81.3 percent were repaired by the communications shop. Conversely, for the work orders in the top 10 percent (the unfavorable errors), shops other than communications repair accounted for 80 percent of the unfavorable errors.
Not only is the relationship between type of repair and type of error interesting, but that relationship is statistically significant. Since it was significant, I took another look at the frequency, discounting any communication repair work orders. The mean now was +.272 man-hours.
This result indicates direction for further research. The clerks may be putting man-hour data into SAMS differently for communications repairs than for the combination of heater, wheeled, and tracked equipment repair. If this error is isolated, the shop officers can eliminate errors causing loss of man-hours.
My evidence suggests that the clerks input man-hours into SAMS on a relatively accurate basis, so maintenance managers should turn their attention to the shop foremens' man-hour accounting. While collecting man-hour data from the work forms, I noticed several inconsistencies that indicated that shop foremen do not record all man-hours.
First, there were work forms on which the shop foremen had recorded no man-hours. Even if the shop foreman had inspected and rejected the work order, he should have recorded some man-hours to account for the inspection. Second, there were work forms showing only initial inspection man-hours, tasks completed, and parts used that indicated more total man-hours than the shop foremen recorded. Third, I observed work order forms with initial, task, and final man-hours recorded, but the total was less than would have been required to complete the repairs.
For instance, someone had recorded a total of 4 hours to change an M728 engine. The Army's maintenance allocation chart (MAC) allows much more than 4 hours to change an engine. However, it might be possible to change an engine in 4 hours if three or four mechanics are working on the job. If four mechanics complete the engine change in 4 hours, the shop foreman should record 16 man-hours to change the engine and at least 1/2-hour each for the initial and final inspections. Instead of recording 4 hours for this job, the shop foreman should have recorded a total of 17 man-hours. This was not an isolated case. I found several instances of this type of error in all the shop records.
Considering these findings, there are several steps maintenance managers should take to make their man-hour accounting systems more effective. I recommend that maintenance managers repeat this research method with a recent sampling of work order forms for their maintenance companies. Repeating the research quarterly using up-to-date samplings will validate the clerks' accuracy. It also gives maintenance managers a current snapshot of their clerks' accuracy and what improvements the clerks have made.
Next, I recommend that the support operations office examine the three areas I've discussed during the next battalion command inspection (BCI) for each of their maintenance companies. To do this, they should examine work orders to see if the shop foremen record the initial inspection man-hours, check to see if work orders show man-hours for tasks completed and parts used beyond the initial inspection, and compare the total man-hours used with the MAC to ensure that total man-hours recorded are comparable to standard repair requirements.
If the support operations office finds accounting errors similar to mine, maintenance managers will have to find and implement corrective actions. These actions possibly could include a better quality control process or man-hour accounting training for shop foremen.
I recommend that the support operations office take action from the lessons learned while inspecting the man-hour accounting process. They can use the man-hour accounting checklist I developed (at left) to help analyze my unit's accounting processes or create their own checklist to use during future inspections.
They also should develop a process that checks the data periodically. By institutionalizing the man-hour accounting inspection, support operations will force shop officers to successfully account for man-hours using the SAMS. The shop officers then can more effectively manage their backlog.
Remember, DO NOT BLAME THE CLERKS! They input man-hour data into SAMS with an acceptable level of accuracy. Although there is always room for improvement, the clerk's accuracy in recording data from the work orders is not significantly in error. Instead of blaming the clerks, shop officers and maintenance managers should study their input data to determine which shop clerks more accurately record data and then determine what caused the errors.
Maintenance managers also must examine the way shop foremen record man-hours. Managers may find that shop foremen inaccurately report man-hours. When shop officers have made appropriate changes to the man-hour accounting process, they finally will be able to reduce maintenance backlog using man-hours that are correctly recorded. ALOG
Captain George C. Martz, Jr., is currently assigned to the 1st Cavalry Division, Fort Hood, Texas. He has held a number of company-level and battalion-level maintenance positions, including support operations maintenance officer, shop officer, and ground support equipment platoon leader. A recent graduate of the Army Logistics Management College's Combined Logistics Officer Advanced Course, he also has completed work for a master's degree in public administration from Golden Gate University, San Francisco, California.