Question 8 of 8: Rebalancing 14 Jobs Across Three Surface-Mount Machines, and Minimum Workforce
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Notes on this paper
National Examinations — May 2019 — 17-Ind-A4 Production Management. Three-hour, closed-book exam; Casio or Sharp approved calculators only. Format: eight questions, each worth 20 marks (10/10 sub-part split per the front-page marking scheme); candidates do two questions from Section A and three from Section B, and only the first five questions appearing in the answer book are marked. All eight are solved below for completeness. The paper asks for point-form answers wherever possible; the solutions below use full working for clarity.
Reference texts: Nahmias & Olsen, Production and Operations Analysis (7th ed., Waveland/McGraw-Hill) — forecasting, inventory (EOQ/EPQ) and aggregate planning; Sipper & Bulfin, Production: Planning, Control, and Integration — production scheduling, JIT/kanban and shop-floor implementation gaps; Hillier & Lieberman, Introduction to Operations Research (11th ed.) — LP formulation and project scheduling (CPM/PERT); Pinedo, Scheduling: Theory, Algorithms, and Systems (5th ed.) — parallel-machine scheduling and days-off workforce scheduling; Hopp & Spearman, Factory Physics (3rd ed.) — variability, buffering, and production scheduling; Liker, The Toyota Way, Shingo, A Revolution in Manufacturing: The SMED System, and Shingo, Zero Quality Control: Source Inspection and the Poka-Yoke System — 5S, Five Whys, poka-yoke, SMED and lean root-cause analysis; R.W. Hall, Zero Inventories — the “seven zeros” JIT framework.
Question 8: Rebalancing 14 Jobs Across Three Surface-Mount Machines, and Minimum Workforce (20 marks)
The 14-job data set (batch sizes and per-machine times) and the workforce shift/days-off rules are as printed (including job B2401’s time of 3,200 s). The target time is stated two ways: this paper prints “complete in 3.5 hours if possible” in the stem, then refers to “the proposed four-hour schedule” in part (b). Part (a) below shows 3.5 h is impossible and the best schedule takes exactly 4 h, which is consistent with part (b).
Given. Fourteen jobs, each with a fixed processing time (seconds) shown once regardless of which machine runs it (the three machines have “similar capabilities,” so a job’s time does not depend on its assigned machine); three identical parallel machines A, B, C; target completion within 3.5 hours ($12{,}600$ s). No individual due dates are given, so every job is taken to share the 3.5-h due date $d=12{,}600$ s. The lateness of job $j$ is $L_j=C_j-d$, so the worst lateness is $L_{max}=C_{max}-d$, and minimizing it is the same as minimizing the makespan $C_{max}$.
Job
Batch size
Time (s)
Initial machine
B2401
72
3,200
A
B7982
126
4,400
A
B6183
45
6,000
B
B1184
110
3,800
A
B9455
240
3,800
C
B4056
32
4,300
B
B1847
32
4,300
B
B6298
32
4,300
B
B9989
192
1,800
C
B1910
64
1,200
B
B3311
64
1,200
B
B8212
32
2,900
B
B4813
64
1,000
B
B7214
64
1,000
B
Initial totals
A 11,400 / B 26,200 / C 5,600
Find. (a) A rebalanced schedule that minimizes the worst lateness against the 3.5-hour (12,600 s) target, and whether that target can be met; (b) the minimum operator headcount needed to staff this facility under the stated shift and days-off rules.
Approach. The initial allocation is badly imbalanced (Machine B alone totals 26,200 s $=7.3$ h, about double the deadline, while C sits at only 5,600 s), so find the theoretical lower bound on makespan, search for a job-to-machine assignment achieving it, then size the workforce from two lower bounds (days off per week, and weekends off per five weeks) and a rota that meets the larger one.
Lower bound. Total work content across all 14 jobs is $\sum_jp_j=43{,}200$ s; with 3 identical parallel machines, no assignment can beat
$$C_{max}\ge\left\lceil\frac{43{,}200}{3}\right\rceil=\boxed{14{,}400\ \text{s}}.$$
This bound is already above the 3.5-h target ($14{,}400>12{,}600$ s), so no schedule can finish all jobs in 3.5 hours. The best possible is to reach the bound, 4 hours, if a perfectly balanced 3-way partition of these 14 job times exists.
Rebalanced assignment (part a). An exhaustive branch-and-bound search over 3-way partitions of the 14 jobs (minimizing the largest machine load) finds a partition that exactly hits the lower bound:
$$\text{A: B6183, B4056, B8212, B1910}\ (6{,}000+4{,}300+2{,}900+1{,}200=14{,}400\text{ s})$$
$$\text{B: B7982, B1184, B2401, B9989, B3311}\ (4{,}400+3{,}800+3{,}200+1{,}800+1{,}200=14{,}400\text{ s})$$
$$\text{C: B1847, B6298, B9455, B4813, B7214}\ (4{,}300+4{,}300+3{,}800+1{,}000+1{,}000=14{,}400\text{ s})$$
giving $C_{max}=14{,}400$ s (4 h 0 min) on all three machines, equal to the lower bound and therefore optimal. The worst lateness against the 3.5-h target is
$$\boxed{L_{max}=14{,}400-12{,}600=1{,}800\ \text{s}\ (30\ \text{min})},$$
the least possible. The job order within a machine does not change $L_{max}$; sequencing shortest-first (SPT) on each machine minimizes how many jobs finish after 3.5 h. This 4-hour schedule is the “proposed four-hour schedule” used in part (b). Method: compute the averaging bound, then search job assignments (largest job first, placed on each machine in turn, pruning any branch whose largest load already reaches the best found) until a partition meets the bound.
Minimum workforce (part b). Three machines, each needing one operator per shift, times three 8-hour shifts/day, seven days/week, gives a constant requirement of $R=3\times3=9$ operators every day (one shift each). Two lower bounds apply:
Days-off bound. Each operator needs two days off in every seven, so works at most 5 of 7 days: $5N\ge7R=63\Rightarrow N\ge\lceil12.6\rceil=13$.
Weekend bound. Every Saturday needs 9 operators, so a 5-week cycle needs $5\times9=45$ Saturday shifts. Each operator must have at least 2 full weekends off in 5 weeks, so works at most 3 of the 5 Saturdays: $3N\ge45\Rightarrow N\ge15$ (Sundays give the same bound).
$$N\ge\max(13,\ 15)=\boxed{15\ \text{operators}}.$$
A 15-person rota that meets it. Form 5 crews of 3. In week $k$ ($k=1\ldots5$), crews $k$ and $k+1$ (cyclically) take the weekend off, so 3 crews (9 operators) work every weekend and each crew gets exactly 2 weekends off in 5. Each crew also rests the same two weekdays every week: crews 1 and 2 Mon+Tue, crew 3 Wed+Thu, crew 4 Thu+Fri, crew 5 Wed+Fri. That leaves exactly 3 crews (9 operators) on duty each weekday. Within a working crew, one operator takes each shift. Because every crew rests a fixed pair of weekdays, any 7 consecutive days contain at least two of its days off, so the rule holds for rolling weeks as well as calendar weeks.
Machine
Job set
Load (s)
A
B6183, B4056, B8212, B1910
14,400
B
B7982, B1184, B2401, B9989, B3311
14,400
C
B1847, B6298, B9455, B4813, B7214
14,400
Makespan and worst lateness (part a)
14,400 s = 4 h on all three machines (optimal); 3.5-h target impossible; $L_{max}=1{,}800$ s (30 min)
Minimum workforce (part b)
15 operators (weekend bound $3N\ge45$ governs; 5 crews of 3)
Figure 3 — Rebalanced assignment: Machines A, B and C each 14,400 s (4 h), the lowest possible makespan; 30 min past the 3.5-h target.