Production planning and scheduling in a high volume injection molding facility.

Arne Thesen

University of Wisconsin

INTRODUCTION

DovrePlast is a high volume producer of injection molded products of various types located in one of the more scenic mountain regions of central Norway. The company sells its products through three different channels:

1. Retail sales through company owned stores.

2. Wholesale sales to other dealers.

3. Contract sales in bulk quantities to manufacturers of other products.

Most products are sold though more than one channel, and the nature of the demand processes and the corresponding delivery arrangements differ significantly from channel to channel (Table 1).

Distribution / Lead time / Demand Process
channel / (time from receipt of order to delivery) / Order Size / Order Intervals / Available information
Retail / Immediate delivery / Random
(1 - 50) / Random / Forecasts
Wholesale / 15 days / Random
(100 - 3,000) / Random / Forecasts
Contract / 1 day / Fixed
(500 - 5000) / Random / Total sales/year,
order size.

Table 1: Demand processes for the three different distribution channels.

Given the tight lead times for retail and contract sales, the company meets these demands by shipping previously produced parts from inventory. On the other hand, wholesale demand is always met shipping products produced after the order has been received. Production is scheduled independently for items destined for the inventory, and items already sold to customers, and delivery delays due to scheduling conflicts are not uncommon. Furthermore, the company feels that it is at a competitive disadvantage as it is unable to accurately predict shipping dates for many items.

A PROBLEM

Recently the company has experienced significant delays in its deliveries of a class of nine high demand products produced in a cell of three dedicated high pressure injection molding machines.

Sales

An analysis of the demand processes for these products were made. The results are shown in Table 2. We note that demand ranges from 860 to 4480 items per week, and that all products are sold through at least two different channels.

Retail / Wholesale / Contract / Total
Parts/Order / Orders/week / Parts/Week / Parts/Order / Orders/ week / Parts/Week / Parts/Order / Orders/week / Parts/Week / Parts/Week
1 / Monitor- assy-1 / 0 / 0 / 0 / 3000 / 1 / 3000 / 4000 / 0.1 / 400 / 3400
2 / Monitor- assy-2 / 0 / 0 / 0 / 3000 / 0.5 / 1500 / 4000 / 0.1 / 400 / 1900
3 / Monitor- assy-3 / 0 / 0 / 0 / 2000 / 0.23 / 460 / 4000 / 0.1 / 400 / 860
4 / Monitor- assy-4 / 0 / 0 / 0 / 2500 / 0.23 / 575 / 4000 / 0.1 / 400 / 975
5 / TV cabinet-18 / 0 / 0 / 0 / 1000 / 0.4 / 400 / 4000 / 0.1 / 400 / 800
6 / TV cabinet-21 / 0 / 0 / 0 / 3000 / 0.4 / 1200 / 4000 / 0.1 / 400 / 1600
7 / TV cabinet-25 / 0 / 0 / 0 / 5000 / 0.4 / 2000 / 4000 / 0.1 / 400 / 2400
8 / Fish Containers / 40 / 12 / 480 / 4000 / 0.9 / 3600 / 4000 / 0.1 / 400 / 4480
9 / File Boxes / 5 / 100 / 500 / 1000 / 2 / 2000 / 0 / 0 / 0 / 2500

Table 2: Demands for nine products produced in a single dedicated cell.

Part routings

The products are produced in a cell consisting of three high pressure injection molding machines (Machines number 146, 148 and 164). All products can be made on machine 146, in addition, most products can also be made on machine 148 or machine 164. Only one product can only be made on machine 146. Since the machines are not identical, cycle times (in seconds per part) for different products differs from machine to machine. On the other hand, setup times depends mostly on the tools (molds) being used, hence setup time for a given product is identical for all machines. Also, parts are produced in the same quantities (batch sized) regardless of the machine used.

‘ / Cycle time(sec) / Setup / Avg.
146 / 148 / 164 / time (hrs) / Batch
1 / Monitor- assy-1 / 72 / 74 / 9999 / 4.5 / 1000
2 / Monitor- assy-2 / 72 / 74 / 9999 / 4.5 / 1000
3 / Monitor- assy-3 / 72 / 74 / 9999 / 4.5 / 1000
4 / Monitor- assy-4 / 72 / 74 / 9999 / 4.5 / 1000
5 / TV cabinet-18 / 65 / 0 / 61 / 4 / 1500
6 / TV cabinet-21 / 65 / 0 / 61 / 4 / 1500
7 / TV cabinet-25 / 65 / 0 / 61 / 4 / 1500
8 / Fish Containers / 65 / 0 / 61 / 4 / 1500
9 / File Boxes / 88 / 0 / 9999 / 4 / 5000

Table 3: Part routings for nine products produced in a single dedicated cell.

ASSIGNMENTS

1. How many shifts must the cell work to meet the demands listed in Table 2 given the part routings shown in Table 3.

2. Find a feasible allocation of parts to machines such that average weekly demands are met. Ignore problems of scheduling delays. Production may be shared between machines in pre determined fractions.

3. Upon inspection of Table 2, one of the firms accountants questioned the wisdom of operating with such small batch sizes. The shop foreman’s response was that “we have always done it this way”, the sales department’s response was “we need small batch sizes to meet our deadlines”. What do you think? Specifically would the company benefit from larger batch sizes, given no other changes in the way products are made/ Support your recommendations with appropriate simulation analysis.

4. The setup times listed above are estimates based on current practice. It is quite likely that these times can be reduced to less that 30 minutes at almost no cost by better planning and improved practices. How would this change affect your answers to question 3?

5. Design a production planning and scheduling system that permits the company to reduce the wholesale lead-times to less than 7 days. (Use the best practices identified above)

Appendix 1: Results from spreadsheet analysis

Machine loading

Machine 146 / Machine 148 / Machine 146
% / Parts / Setups / Set up hrs / Run hrs / % / Parts / Set ups / Set up hrs / Run hrs / % / Parts / Setups / Setup hrs / Run hrs / Total %
1 / Monitor- assy-1 / 0 / 0 / 0.0 / 0.0 / 0.0 / 100 / 3400 / 3.4 / 15.3 / 69.9 / 0 / 0 / 0.0 / 0.0 / 0.0 / 100
2 / Monitor- assy-2 / 0 / 0 / 0.0 / 0.0 / 0.0 / 100 / 1900 / 1.9 / 8.6 / 39.1 / 0 / 0 / 0.0 / 0.0 / 0.0 / 100
3 / Monitor- assy-3 / 45 / 387 / 0.4 / 1.7 / 7.7 / 55 / 473 / 0.5 / 2.1 / 9.7 / 0 / 0 / 0.0 / 0.0 / 0.0 / 100
4 / Monitor- assy-4 / 100 / 975 / 1.0 / 4.4 / 19.5 / 0 / 0 / 0.0 / 0.0 / 0.0 / 0 / 0 / 0.0 / 0.0 / 0.0 / 100
5 / TV cabinet-18 / 60 / 480 / 0.3 / 1.3 / 8.7 / 0 / 0 / 0.0 / 0.0 / 0.0 / 40 / 320 / 0.2 / 0.9 / 5.8 / 100
6 / TV cabinet-21 / 0 / 0 / 0.0 / 0.0 / 0.0 / 0 / 0 / 0.0 / 0.0 / 0.0 / 100 / 1600 / 1.1 / 4.3 / 28.9 / 100
7 / TV cabinet-25 / 78 / 1872 / 1.2 / 5.0 / 33.8 / 0 / 0 / 0.0 / 0.0 / 0.0 / 22 / 528 / 0.4 / 1.4 / 9.5 / 100
8 / Fish Containers / 0 / 0 / 0.0 / 0.0 / 0.0 / 0 / 0 / 0.0 / 0.0 / 0.0 / 100 / 4480 / 3.0 / 11.9 / 80.9 / 100
9 / File Boxes / 100 / 2500 / 0.5 / 2.0 / 61.1 / 0 / 0 / 0.0 / 0.0 / 0.0 / 0 / 0 / 0.0 / 0.0 / 0.0 / 100
TOTAL / 3.4 / 14.4 / 131 / 5.8 / 26.0 / 118.7 / 4.6 / 18.5 / 125.1
145 / 145 / 144 / 433
Avg. Retail / Avg. Wholesale / Avg. Contract
Parts/Order / Order Interval (hrs) / Parts/ Order / Order Interval (hrs) / Parts/ Order / Order Interval (hrs)
1 / Monitor- assy-1 / 0 / inf / 3000 / 168.0 / 4000 / 1680
2 / Monitor- assy-2 / 0 / inf / 3000 / 336.0 / 4000 / 1680
3 / Monitor- assy-3 / 0 / inf / 2000 / 730.4 / 4000 / 1680
4 / Monitor- assy-4 / 0 / inf / 2500 / 730.4 / 4000 / 1680
5 / TV cabinet-18 / 0 / inf / 1000 / 420.0 / 4000 / 1680
6 / TV cabinet-21 / 0 / inf / 3000 / 420.0 / 4000 / 1680
7 / TV cabinet-25 / 0 / inf / 5000 / 420.0 / 4000 / 1680
8 / Fish Containers / 40 / 14 / 4000 / 186.7 / 4000 / 1680
9 / File Boxes / 5 / 1.68 / 1000 / 84.0 / 0 / inf

Appendix 2

A simplified model for research

Types of demands

8 products

equal sales for all products

No retail sales

Equal mix of contract and wholesale demands

Interval between demands = e(1 day)

Contract sales

All contract salesfor 3,000 units

Shipment due 5 days after receipt of order

Wholesale demands

demand = 750 + 500* index ( i.e.1250 for type1 ... to 4750 for type 8, avg = 3,000)

Shipment due 15 days after receipt of order

Part routing

Cycle Times = 90

Parts 1 through 4 made on machines 1 and 2

Parts 5 through 8 can be made on machines 1 and 3

The problem

Find a production schedule in real time that minimizes

Fraction of late shipments

Average inventory levels