Sun Products Company Sun Products Company (SPC) makes a range of components for various customers in the automotive industry at home and abroad. One of these customers made an approach year or so ago with a quality concern about the sunroof which was sup- plied by SPC. The customer delegation told the Sun QA Manager, Fred Jarvis: 'This is on our top 40 list of quality concerns at present. The height of the glass on the roof is often noticeably uneven, or sits too proud of the car body. We'd like to take a look around your processes to see if there are any immediate improvements we can make. After looking round the sunroof line at SPC, the customer delegation suggested to Fred that some improvements should be made to materials handling within the Sun works, and in the way that parts were transported from Sun suppliers. A detailed inspection of the production process with Fred and his colleagues had been made, and a process flow chart drawn up. But attention had eventually focused on the profile set- ting operation where the glass was fitted into the sunroof mechanism. A sketch of this is shown in Figure 44.1. A 100 per cent inspection of the profile was made by probes at the end of the production process. The probes measured the height of the glass at four points on the sunroof near to the corners, represented by P1, P2, P3 and P4. The customer specification called for a profile measure with a tolerance of 1.00mm for each point P1 to P4. If the probes detected a condition which was out of specification for one of these points, the inspector called for the profile setting to be altered. A month later, the customer delegation returned to Fred, and told him that they had made further investigations of their own assembly process. "We have made a number of immediate improvements on our assembly line, such as giving the operator a step block to make the fitting process more exact and less visual. Now we want you to introduce variable control charts on your profile setting operation. You are generating a lot of data at the inspection stage, but you're not doing enough with it! Further, you're reacting to events rather than controlling them.' These comments had come as rather a shock to Fred, who thought that Sun had installed the best kit, and that it was virtually impossible (given 100 per cent inspec tion of the profile) for defective products to be shipped. Nevertheless, Fred got to Figure 44.1 The sunroof at Sun Products Company xp2 P1 P3 P4 (NB: Sketch only: not to scale!) Glass Motor T rilldogs work, first tackling a process capability study. This came as an even greater shock! The process was clearly out of control. It seemed very difficult to maintain specifica- tion on all four points at the same time. Thus Cpk values of less than 1 were initially quite common. Fred concentrated first on getting rid of special causes of variability. For example, he found that when the glass supplier changed batches, the curvature of the glass could change significantly and cause problems at Sun. But once these problems had been recognised and dealt with, Fred was satisfied that control chart- ing could start. Although the inspection process would be left at 100 per cent for the time being, Fred decided on taking two samples each hour and recording the difference to spec- ification for all four points P1 to P4. Samples would be taken for the two readings made closest to the hour. Tables 44.1 and 44.2 show the data which were used for calculating the grand average, and the upper and lower control limits for the aver- ages chart (X Bar Chart) for point P3. They also provide the data used to calculate the upper control limit for the range chart (R Chart) for P3. Table 44.1 P3 data used to calculate the process capability indices Table 44.2 P3 data from a subsequent day's production Reading 1 Reading 2 Reading 1 Reading Saturday 2/12 06.00hrs +0.6mm +0.2mm Thursday 18/1 06.00 -0.3 -0.1 07.00 +0.3 +0.2 07.00 -0.4 -0.1 08.00 +0.4 +0.1 08.00 -0.2 -0.3 09.00 +0.1 -0.3 09.00 -0.2 -0.2 10.00 -0.2 +0.0 10.00 -0.5 -0.2 11.00 +0.0 +0.0 11.00 -0.4 -0.8 Monday 4/12 06.00 +0.2 12.00 -0.7 -0.3 +0.2 13.00 +0.1 -0.1 07.00 -0.3 +0.1 14.00 -0.1 -0.3 08.00 -0.3 +0.0 15.00 -0.3 -0.1 09.00 +0.1 +0.3 16.00 +0.1 +0.2 10.00 +0.6 +0.4 17.00 +0.1 -0.1 11.00 +0.0 -0.2 18.00 -0.2 -0.1 12.00 +0.5 +0.0 19.00 -0.1 +0.2 13.00 +0.1 +0.2 14.00 +0.1 +0.0 Friday 19/1 06.00 -0.6 -0.4 15.00 +0.2 +0.3 07.00 +0.0 -0.4 16.00 +0.4 +0.3 08.00 -0.3 -0.8 17.00 +0.1 -0.1 09.00 -0.5 -0.2 10.00 -0.5 -0.2 Tuesday 5/12 06.00 +0.0 +0.6 11.00 -0.5 -0.2 07.00 +0.2 +0.0 12.00 +0.0 +0.0 08.00 -0.3 -0.1 13.00 +0.1 -0.2 09.00 +0.0 -0.2 14.00 +0.2 -0.1 10.00 +0.5 +0.2 15.00 +0.0 +0.6 11.00 +0.1 -0.2 16.00 +0.2 -0.1 12.00 -0.2 -0.1 17.00 +0.2 -0.1 13.00 +0.0 +0.1 18.00 +0.0 +0.0 14.00 +0.1 +0.1 15.00 +0.0 +0.4 16.00 -0.3 -0.1