Control Charts for Investigation Defects in Production Process - Essay Example

Control Charts for Investigation Defects in Production Process Background Green Field Limited is a small scale enterprise that specializes in the production of clean fuel for household usage. Following the company’s participation in an international expo on clean fuels in Copenhagen and the selection of its Banda Bricks, an international green organization has selected the company to benefit from its Enterprise for Safer Future Award. The award consists of financial and technical support for selected enterprises as part of the organization’s campaign for checking climate change. Green Field started production in 2007. It was set up by two university professors, who invented the Banta Brick technology. The duo started the enterprise to experiment their laboratory findings in the market but they were overwhelmed by the acceptance of their products. Given the increasing demand for Banta Bricks, the management of the company has been discussing plans to transform the company from a mere experimental enterprise to money making venture. Thus the award is an opportunity for the business to grow. However, this does not require only increase in production but consistency in production. The capability of the production process must be determined to enable the company set achievable goals as it increases production. Purpose of Investigation The records of production in the first two years have indicated abrupt fall in production. To know the causes of this abruptness in production and how to correct it the management of the company need to carry out an investigation. The best way to do this is to task the manager in charge of production. To validate the cursory observations, the manager should subject the production data to statistical analysis. She should use control charts in her investigation to identify undesirable trends, causes of the abrupt fall in production and how to improve the process performance so as to achieve process stability. The manager has the following questions to answer on the production process: 1) Are the observed fall in production verifiable? 2) What are the causes of the fall in production? 3) How can such anomalies be prevented and the production process improved? Using Control Charts to Investigate Defects in Production Process Control charts are statistical tools used to establish whether variations in a process are as a result of common causes or special causes. Generally, control charts are used in the i) monitoring of variations in processes, ii) differentiating causes of identifies variations, iii) assessing effectiveness of changes in processes and iv) communicating performance of process (http://www.balancedscorecard.org/Portals/0/PDF/control.pdf). By so doing, control charts help in achieving and maintaining process improvement and stability. Depending on the type of data (variable or attribute) different types of control charts are used to establish statistical control. The commonest charts used include X-chart, R-chart, X-bar and S-chart, U-chart, MA-chart, P-chart, C-chart and NP-chart (http://www.balancedscorecard.org/Portals/0/PDF/control.pdf and Burk 2006). Where variations are established to be caused by special causes (factors not inherent in the process), certain graphical tools are used to investigate these causes. A commonly used tool is the Fishbone, which relies on brainstorming. The fishbone diagram, for example, helps analysts to identify, sort, display and analyze possible causes and effects of variations in processes (ILO 2008: 1). To accomplish the task assigned to her, the manager will need the services of a data analyst. The data analyst will randomly sample five out of the company’s seventeen factory units and use the monthly production for the analysis. Each factory unit has one production machine which can be operated by individual workers. Some useful production statistics are shown in Box 1. Box 1: Production Statistics Number of Machines 17 Number of Factory Workers 54 Number of Factory Worker/per Machine 3.17 Average Man-hours 11 Normal Production Capacity 1/5 per hour Average Actual Production of Machine/Day 45 bundles per day Assuming the sample size is 5 and type of data is variables data, the data analyst should use X-chart and R-chart in his investigation. The unit production (in bundles of bricks) of the five units is shown in Box 2. Using this data the averages (x) of monthly production of the five factory units, ranges (R), Upper Control Limits for monthly production averages (UCLx), Lower Control Limit for monthly production averages (LCLx) and Upper Control Limit for ranges of monthly production (UCLR) will be computed using Microsoft Excel. The computer processed parameters are shown in Box 2. Two constants (A2=0.577 and D4=2.144) for the sample size (n=5) where used to calculate the upper and lower control limits for the monthly production averages. There is no lower control limit for ranges since the sample size is less than seven (Basic Tools, n.d., 13). The monthly averages, ranges of monthly production, and the upper and lower control limits for monthly production averages and upper control limit for ranges of monthly production were plotted on a graph. The X-chart and R chart are shown in Box 3.             Box 2: Monthly Production of Factory Units (Bundles)           Jan Feb Mar April May June July Aug Sept Oct Nov Dec     1 1,645 1,587 1603 1599 1203 1595 1480 1566 1597 1009 1568 1698     2 1706 1451 1609 1546 1199 1501 1478 1490 1579 1111 1598 1703     3 1388 1601 1602 1691 1254 1597 1505 1543 1587 1097 1493 1678     4 1698 1586 1390 1457 1003 1332 987 1001 1134 1003 1521 1699     5 1675 1597 1609 1692 1120 1700 1543 1698 1602 1122 1276 1701     Average 1,622 1,564 1563 1597 1156 1545 1399 1460 1500 1068 1491 1696 1,472     Ranges 318 150 219 235 251 697 556 697 468 119 322 25 338     Upper Control Limit (Averages) 1667     Lower Control Limit (Averages) 1277     Upper Control Limit (Ranges) 714.7                                   The control charts show some anomalies in monthly averages. The averages for May and October are below the lower control limit while the average for December is above the upper control limit. These can be considered as abnormal deviations or defects in the production process. These defects suggest that the Banta Bricks production process is not mature and it is not entirely under statistical control. It implies that the causes of the defects are not typical of the production process. The defects results from special causes, which need to be investigated. It should also be noted that there is no anomaly in the R-chart. All the ranges of monthly production plotted fall within the upper control limit. This means that the differences in production for the five factory units are normal and under statistical control. Box 3: X Chart and R Chart for Banta Bricks Production Process       Averages              Averages                                                         Ranges                                                     Legend           Upper Control Limit       Grand Mean       Lower Control Limit                       The Result of Root Causes Analysis Using Fishbone As it is shown in the control chart, not all the monthly averages fall within the upper and lower control limits in the X-chart. The averages for December went above the upper control limit while those of May and October fell below the lower control limit. This confirms the fall in production observed. The causes of these falls will have to be investigated. To do this the manager should use the brainstorming tool to identify possible causes, which can be diagrammatically captured in a Fishbone diagram as illustrated below (Fig 1). To carryout the investigation, the Production Manager should interview and brainstorm the with factory workers on the possible causes of the observed problem. The causes of the sudden drop in production observed in the control chart will be identified by so doing. The fall in production in May was as a result of low supply of raw materials, machine breakdown and poor power supply. In October the warning strike embarked by factory staff over wage increase coupled with poor power supply and breakdown of power generating set brought down production to its lowest in the year. Also, as could be seen in the production data, break down of machine contributed to the below-average productions in some factory units particularly factory unit 4. The low production in this is considered normal since it falls within the control limits. The machines used for the production of the bricks are substandard. They are not produced to handle production for more than 6 hours. And because wages are tied to the quantity produced by factory units, factory workers work for an average of 11 hours per day including weekends. There is no shifting arrangement as individual workers compete to produce more using the same factory machines. The machines are therefore over stretched as the produce over 250% of the production capacity on a daily basis. The organizational culture of ‘earn as you produce’ adopted by the company is not helping matters. It does not encourage teamwork as individuals struggle to use fewer machines to produce competitively. This has resulted in frictional relationship among factor workers, which often leads to physical fighting and stalling of production. In addition, the factory workers were not adequately trained on the use and maintenance of machines. They were recruited based on their experiences in handling medium factory plants. The manuals for operating the machines are English hence not digestible to workers, who are largely high-school drop outs. The machines used in the factory are not locally fabricated and their spare parts are not readily available in the market. This complicates maintenance and unnecessarily elongated period of repairs. Another factor responsible for the fall in production is epileptic power supply and over-working of the power generating set. The factory gets an average of two and half hours supply of electric power in a day. This makes the factory to rely on its 125 KVA generating set, which is produced to work for not more than seven hours. However, it powers all the seventeen production machines. Due to the over stretched use of the generating set and couple with the lack of maintenance; it broke down twice in the year reviewed. The set was not new when it was installed in the factory. It had less than 24 months of its premium output span remaining when it was installed. The raw materials for the production of Banta bricks are household and farm wastes. Household waste accounts for only twenty five percent of the raw materials used in the production. This is largely due to the poor waste management practice in the city. Therefore the factory relies largely on agricultural waste from farms. Unfortunately, agriculture is largely rain-fed and seasonal. It is therefore exposed to the vagaries of the tropical climate, which is highly unreliable. Short spells or drought can cause significant drop in agricultural production. This leaves the factory with little supplies from the farm as it competes with farmers and cattle for the farm waste. This was partly the cause of the drop in production in May. There was drought in the preceding year, which compelled farmers to use most of the waste as domestic fuel leaving the remaining for farm animals and cattle of pastoralists. Fig 1: Fishbone Diagram Showing Causes of Fall in Banta Bricks Production Process Substandard Over-worked Power Machines Generating Set Seasonal raw materials Epileptic power supply No spare parts Teamwork discouraged Inadequate training Poor understanding Production-based Strike of Manuals Payment Finally, the warning strike action embarked by the factory workers was largely responsible for the drop in production in October. The workers began agitating for wage increase and welfare package given the increase in production and profits. Failure of the management of the company to respond and the factory accident that claimed an arm of a worker led to a warning strike action. Production was totally stopped for one week. The company exhausted products in the stores. This compelled the company to review the wages of workers and started an insurance policy in case of factory accidents. Bonuses and one-week leave grant were also initiated. This resulted in the increase in sudden increase of production after the strike and the exceptional production in December. The management promised 50% of the total wage of workers as Christmas bonus and one week holiday to celebrate the festive period. References Basic Tools For Process Improvement: Module 10 Control Charts. Web. March 28, 2010. http://www.balancedscorecard.org/Portals/0/PDF/control.pdf. Burk, S, 2006, A Better Statistical Method for A/B Testing in Marketing Campaigns. Marketing Bulletin, Technical Note 3. Web. March 30, 2010. http://marketing-bulletin.massey.ac.nz Donald J. Wheeler, 1995, Advanced Topics in Statistical Process Control, SPC Press International Labour Organization, 2008. Factory Practice Guide: Using Fishbone Diagram as a Problem Solving Tools, Web. April 4, 2010. www.ilo.org/factoryimprovementprogramme Read More