Introduction:
Donner Company, a customized printed circuit board manufacturer is facing three main problems: productivity, quality issues and delayed deliveries. We start analyzing with a diagram showing information flow beginning from the customer placing the order to the delivery of that order can be represented as below.
Customer places order
Customer places order
Estimate Labour and Material costs
Estimate Labour and Material costs

Prepare and submit bid
Prepare and submit bid
Check & Verify Design
Check & Verify Design

Detailed specifications and factory order by Altmeyer
Detailed specifications and factory order by Altmeyer
Normal
Normal
Customer accepts bid?
Customer accepts bid?
Yes
Yes
Promise delivery time
Promise delivery time

1 day from bid accept-ance
1 day from bid accept-ance
Rush
Rush
Scheduling & processing of order
Scheduling & processing of order
4 days from bid accept-ance
4 days from bid accept-ance
2 days
2 days
No
No
End
End
Material specifications by Schnabs
Material specifications by Schnabs
Delivered to customer
Delivered to customer
Send order to purchasing agent
Send order to purchasing agent
Shipping the order
Shipping the order
Instruct purchasing agent & procure material
Instruct purchasing agent & procure material
Blue print, factory order sent to Flaherty
Blue print, factory order sent to Flaherty
Blue print, factory order sent to Flaherty
Blue print, factory order sent to Flaherty
Order Raw material
Order Raw material

Capacity Analysis:
Since Donner Company manufactures PCBs according to the specifications of various customers, the PCBs being manufactured are not of the same type. The company can be considered as a low-volume, high-variety product manufacturer. So, each order does not go through the same sequence of processes and also does not take the same amount of time. Assumptions: 1. Even though it is mentioned that some orders may follow different sequence of process and may even skip some of the processes, we assume that every order goes through each process mentioned in Exhibit 2. 2. Proportion of order size and the no. of orders will be the same in future months; same as September given in Exhibit 4. 3. Each order requires separate setting up of equipment. So, when the next order arrives, the equipment must be setup once again in each process (artwork, inspect, etc.)
The time required in each process for different size orders will not be the same because of the presence of two components-setup time (common for one order) and run time (varies with the number of boards in each order). As the order size increases, the time spent in the setup time is shared among all the circuit boards and hence time required per board is reduced.
The amount of time required to be spent per order in each of the processes is calculated for each of the order sizes from 1 to 1050 as given in Exhibit 4. For all order sizes, the drilling process (manual drilling up to order size 5 and CNC drill for order size above 5) takes the maximum amount of time per order. So, the drilling process is the bottleneck in the system. We assume that only one of the seven drill presses is used. The time taken in drilling process per order is shown in Appendix A. The time taken per order for manual drilling is less than the time taken for CNC drilling up to order size of 5. Hence, manual drilling should be done if order size ≤ 5 and CNC drilling otherwise. The bottleneck time is taken accordingly. As can be seen from the table, different order sizes lead to different capacity for the plant. If the order size which the company receives is always 1, the company has the capacity to manufacture 8.7 PCBs/day. However, if the order size is always 1050, the company can produce 215.4 PCBs per day. This value is arrived at considering that only one of the seven drilling presses is used in the case of manual drilling. Since different order sizes are mixed and considering the same proportion of order size mix will be followed in the future, the average capacity of the plant is found out by calculating the weighted average of capacities for different order sizes. The weightage given is the no. of boards produced per month belonging to that particular order size to the total no. of boards produced per month. In this way, the average capacity of the plant is found out to be 180.36 boards/day.
CNC Drill
Assuming order size is ‘n’, we have to find the value of n where the rate of producing of CNC drill is greater than manual drill. For the manual drill, setup time is 15 mins, each hole takes 0.08 mins and each board has 500 holes, hence the equation for total time taken can be written as 15+0.08*500*n. Similarly, for the CNC drill setup time is 240 mins, each hole takes 0.004 mins and each board has 500 holes, hence the equation for total time taken can be written as 240+0.004*500*n. Now we have to solve the following equation to arrive at n above which CNC should be used.
240+0.004*500*n≥15+0.08*500*n
n≥5.92
We round off n to the nearest natural number 6. Hence, we propose that if n≥6 use CNC drill. In the above calculations we assumed that only one drill is being used at a time. Now we repeat the calculations for 7 drills which gives an n value of 61.
240+(0.004*500*n)7≥15+0.08*500*n
n≥61

CNC Router
Proceeding in a similar way as we did above for CNC drill, the punch press setup time is 50 mins, each circuit board takes 1 minute, and hence the equation for total time taken can be written as50+0.5*n. Similarly, for the CNC router setup time is 150 mins, 1 board takes 0.5 mins, hence the equation for total time taken can be written as 150+0.5*n. If n≥200 use CNC router
50+n≥150+0.5*n
n≥200
Standard Labor time The amount of time to be spent in each of the processes from artwork generation to inspect, test, pack is calculated and added to find the standard labor time for an order size of 1. Similar approach is followed for all the other order sizes as well. For example, in the case of image & shear, from Exhibit 2, setup time per order = 20 minutes and run time = 0.5 minutes per panel Each panel consists of 8 circuit boards so, total time per order (order size1 = 1 PCB) = 20 + (0.5*order size/8) = 20. 0625 minutes. Calculating time in each of the processes in this way, standard labor time for an order of 1 board = 383.38 minutes. Similarly, Standard labour time for 8-board order = 719.05 mins. Standard labor time for an order size of 200 boards = 1840.25 mins. For order size of 8 & 200, we assumed that CNC drill is used, but in case of order size of 1 board, we use manual drilling press. Punch press is used for profiling in all 3 cases. For order size of 200, profiling with either of punch press & CNC router will take the same time. It is observed that since August, the customer returns has increased from under one percent to three percent and the shipments had averaged nine days late. From the profit and loss summary shown in Exhibit 1, it is observed that the net sales has jumped from $85700 in July 1987 to $144400 in August 1987. This sudden increase in sales must have increased the amount of work required. But Donner Company was following the same process as before. This might have caused some pressure on workers, increased delay time and deteriorated the quality. It is also mentioned that nine-tenth of the boards being returned are because Donner had missed or failed to complete one or two of the required operations. This clearly shows that completing all the operations within time has been an issue. But to tackle the increased sales in August, Plummer had employed 8 new people. Their skill is not as good as the old employees. But assuming that they are as skilled as old employees, a comparison of labor utilization rate for the months of July, August and September is done in Appendix B Net sales in September = $124836 No. of boards shipped = 5761 Value of shipment of one board = 124836/5761 = $21.67 Assuming value of shipment per board to be constant, September data | week1 | week2 | week3 | week4 | | 1 to 7 | 8 to 14 | 15 to 21 | 22 to 29 | Value of Average delivery per week($) | 17909 | 18627 | 32652 | 48692 | Value of shipment per board($) | 21.67 | 21.67 | 21.67 | 21.67 | Average boards shipped per week | 826 | 860 | 1507 | 2247 | Labour time required(mins) | 9451 | 9830 | 17231 | 25696 | Labour utilization rate | 14% | 15% | 26% | 39% | The no. of employees was considered as 22 in July and 30 in August and September. The fact that 4 employees (who serve as assistant supervisors) spend only 90% of their time in actual production is also considered in calculating time spent in plant by all employees. We observed that labor utilization rate increased from 22% in July to 27% in August in spite of adding 8 employees. The average processing time increases with increase in labor utilization rate. Even though labor utilization rate has only marginally increased from 22% to 27%, the fact that the new employees are not very productive adds on to this effect and led to delivery time delay. But this does not explain the entire problem. It is mentioned that Plummer emphasizes a shipping policy aimed at clearing all the work possible out of the shop prior to the end of each month. This practice might cause low labor utilization in the initial part of the month and high labour utilization in the later part. Considering the data given in Exhibit 5, one month is divided into 4 weeks and amount of shipments is divided among the four weeks. Labor utilization rate in each of the four weeks is calculated.
From the table, labor utilization rate keeps increasing with every week. So, by the end of the week, when most of the shipment is taking place, the utilization rate is nearly 39% which means the average processing time will be more than normal. Hence, the time taken to complete the orders will be more and this explains the increase in delay period and also the increase in customer returns. Hence, the following are some of the specific actions to be taken to address the problems. 1. The shipping policy needs to be modified. The no. of shipments should be spread evenly over the months so that nearly constant labor utilization rate is maintained. This will decrease the average processing time and also improve the customer returns. 2. It is also mentioned that only orders for more than 100 boards should be drilled in CNC drill. The amount of time spent in the drilling process is high because of this policy in the case of orders sizes 6 to 100. Instead of following this policy, order sizes above 5 should be drilled using CNC drill and others should be drilled using manual presses.

Appendix A Order size | Drill manual(mins) | CNC drill(mins) | Bottleneck=Time per order(mins) | Cycle time per board=time per board/order size | Capacity (units/day) | 1 | 55 | 242 | 55 | 55.0 | 8.7 | 2 | 95 | 244 | 95 | 47.5 | 10.1 | 3 | 135 | 246 | 135 | 45.0 | 10.7 | 4 | 175 | 248 | 175 | 43.8 | 11.0 | 5 | 215 | 250 | 215 | 43.0 | 11.2 | 6 | 255 | 252 | 252 | 42.0 | 11.4 | 10 | 415 | 260 | 260 | 26.0 | 18.5 | 11 | 455 | 262 | 262 | 23.8 | 20.2 | 12 | 495 | 264 | 264 | 22.0 | 21.8 | 14 | 575 | 268 | 268 | 19.1 | 25.1 | 20 | 815 | 280 | 280 | 14.0 | 34.3 | 40 | 1615 | 320 | 320 | 8.0 | 60.0 | 50 | 2015 | 340 | 340 | 6.8 | 70.6 | 60 | 2415 | 360 | 360 | 6.0 | 80.0 | 84 | 3375 | 408 | 408 | 4.9 | 98.8 | 100 | 4015 | 440 | 440 | 4.4 | 109.1 | 113 | 4535 | 466 | 466 | 4.1 | 116.4 | 136 | 5455 | 512 | 512 | 3.8 | 127.5 | 140 | 5615 | 520 | 520 | 3.7 | 129.2 | 154 | 6175 | 548 | 548 | 3.6 | 134.9 | 200 | 8015 | 640 | 640 | 3.2 | 150.0 | 229 | 9175 | 698 | 698 | 3.0 | 157.5 | 252 | 10095 | 744 | 744 | 3.0 | 162.6 | 800 | 32015 | 1840 | 1840 | 2.3 | 208.7 | 1000 | 40015 | 2240 | 2240 | 2.2 | 214.3 | 1050 | 42015 | 2340 | 2340 | 2.2 | 215.4 |

Appendix B | July | August | September | Sales ($) | 85700 | 144400 | 124800 | Value of shipment per board | 21.67 | 21.67 | 21.669 | No. of boards shipped per month | 3955 | 6663 | 5759 | Time required for producing the boards (mins) A | 45228 | 76207 | 65863 | Time spent in plant by all employees (mins) B | 207360 | 284160 | 284160 | Labour utilization rate = A/B | 22% | 27% | 23% |

The worksheet used for various calculations is attached below