Ch. 22 & 22S — Quick Review Q&A

Ch. 22 — Workcenter Scheduling

1 What exactly is a "workcenter," and why does scheduling one matter?

A workcenter is any area where productive resources are organized and work gets done. It could be a machine, a desk, a service counter — any resource that processes jobs.

🍕 Campus Example: Think of the different stations in your campus dining hall — the grill station, the salad bar, the pizza oven. Each is a workcenter. Scheduling decides which meal orders get made in what sequence, on which station, and when.

The goal of workcenter scheduling is to allocate jobs to resources, sequence them, dispatch (release) them, and monitor status — all while minimizing work-in-process (WIP) inventory.

2 Why can't we just use "first come, first served" for everything?

FCFS feels "fair," but it often produces the worst average flow time and the most WIP. If a 2-minute job gets stuck behind a 45-minute job, everyone downstream waits.

🧺 Campus Example: Imagine the dorm laundry room has one washer free. Someone ahead of you has 3 huge loads (90 min total). You just need to wash one small bag of gym clothes (20 min). Under FCFS, you wait 90 minutes before you even start. Under SOT, you'd go first, dramatically reducing average wait time for everyone.

This is why operations managers use priority rules — simple decision rules that use one or two pieces of information about the jobs to determine order.

3 What are the major priority rules, and when does each shine?

Rule	Logic	Best For
FCFS	Run jobs in arrival order	Service fairness (bank lines, DMV)
SOT / SPT	Shortest processing time first	Minimizing avg. flow time & WIP
EDD	Earliest due date first	Minimizing maximum lateness
STR	Smallest slack time remaining first	Urgency-aware scheduling
CR	Critical ratio (time left ÷ work left)	Dynamic, recalculated often
LCFS	Last come, first served	Happens by default (stack of papers)

STR = Time remaining before due date − Remaining processing time
CR = (Due date − Current date) / Work days remaining

📚 Campus Example: It's finals week and you have 5 exams. SOT says: knock out the easy ones first (short study time) to free up mental bandwidth. EDD says: study for whatever exam is earliest. STR says: figure out which exam has the least cushion between "time left" and "study hours needed" and prioritize that one.

Research shows SOT is optimal for minimizing average flow time — but the best practical rules should be dynamic (recalculated frequently) and based on slack.

4 If SOT is "optimal," what's its fatal flaw?

SOT constantly pushes long jobs to the back of the line. In a dynamic shop where new short jobs keep arriving, a big job can be indefinitely postponed.

🎟️ Campus Example: At a concert merch table, if the vendor always serves the quickest transaction first (t-shirt purchases) over someone trying to buy a custom bundle, that bundle customer might never get served. You need a "lateness" safety valve: once a job has waited too long, bump its priority up regardless of processing time.

A bank manager wouldn't use SOT either — customers can't tell in advance how long their transaction takes, and the perceived unfairness of cutting the line would be terrible for customer satisfaction.

5 How does Johnson's Rule optimize a two-machine sequence?

When every job must pass through Machine A then Machine B (in that order), Johnson's Rule minimizes total completion time:

Step 1: Find the shortest processing time across both machines. Step 2: If it's on Machine A, schedule that job as early as possible. If it's on Machine B, schedule it as late as possible. Step 3: Remove that job and repeat.

🚚 Campus Example: A food truck has two steps: Prep (chopping, mixing) → Cook (grill/fry). You have 7 different orders. Johnson's Rule tells you how to sequence them so the total time from first prep to last plate is minimized — keeping both the prep person and the cook busy with minimal idle gaps.

6 When and why do we use the Assignment Method?

The assignment method is a special case of linear programming used when you have n things to distribute to n destinations, each assigned to exactly one, using a single criterion (like cost or time).

🎓 Campus Example: The tutoring center has 4 tutors and 4 subjects to cover tonight. Each tutor has different proficiency ratings per subject. The assignment method finds the one-to-one matching that maximizes total proficiency (or minimizes total cost).

7 What role do Gantt Charts play in shop-floor control?

Gantt charts give a visual timeline showing which jobs are on which machines, when they're scheduled to start/finish, and actual progress versus the plan. They work for both project management and shop-floor control because both involve scheduling multiple tasks across finite resources over time.

📋 Campus Example: Picture a campus event coordinator managing three venues for homecoming weekend. A Gantt chart shows which event is in which venue, when setup/teardown happens, and whether anything is running behind. The "point in time" marker shows where you are right now versus the plan.

The difference from project Gantt charts: in shop-floor control, jobs usually aren't interdependent — their precedence is set by the scheduler's priority rules, not by task dependencies.

8 Why is service-sector scheduling (like staffing) so much harder than manufacturing scheduling?

In manufacturing, you control the input rate. In services, customers arrive when they want and expect service immediately. You can't "hold" a rush of customers and smooth them out over the day.

☕ Campus Example: The campus coffee shop gets slammed from 8:15–9:00 AM between classes. If you staff enough baristas for the rush, they're idle from 9:00–11:00. If you staff for average demand, the rush creates 15-minute lines. Balancing labor cost against customer satisfaction is the core tension. A Service Execution System (SES) helps link, schedule, and track these customer encounters.

Predictable demand patterns help (class schedules, lunch hours), but workers expect a minimum shift length once they arrive — so you often have excess capacity during slow periods.

Ch. 22S — Theory of Constraints (TOC)

9 Why is "perfectly balanced capacity" actually a trap?

Goldratt's key insight: in a perfectly balanced system, natural variability means any station that falls behind has no way to catch up. Small disruptions cascade, causing missed deadlines and growing WIP. The better strategy is to balance the flow of product, not the capacity of each station.

🚶 Campus Example: Imagine 5 friends walking single-file to class. If everyone walks at exactly the same speed, one person tying a shoe creates a growing gap that nobody can close. But if the people behind the slowest walker have extra speed capacity, they can catch up after any disruption. That's why unbalanced capacity (with slack in non-bottleneck resources) actually keeps the system flowing.

10 What's a bottleneck vs. a CCR — and why does every lost minute at a bottleneck matter?

A bottleneck is any resource where capacity is less than the demand placed on it (utilization > 100%). A capacity-constrained resource (CCR) is close to demand but not quite over — it can become a bottleneck if not managed carefully.

The critical principle: an hour lost at a bottleneck is an hour lost for the entire system. An hour saved at a non-bottleneck saves nothing — it just creates idle time.

✈️ Campus Example: During move-in day, the single freight elevator is the bottleneck. If it breaks for 30 minutes, that's 30 minutes of throughput the entire building can never recover. Meanwhile, the lobby check-in desk (a non-bottleneck) being idle for 30 minutes costs nothing — there's nobody to check in faster than the elevator can move them.

Utilization = Demand on Resource ÷ Available Capacity
If utilization > 100% → Bottleneck

11 Explain Drum-Buffer-Rope like you're explaining it to a friend.

Drum = the bottleneck. It sets the pace (the "beat") for the entire system. Buffer = a safety stock of inventory placed just before the drum so it never starves. Rope = communication back upstream telling the first station to only release material at the rate the drum can process it.

🎶 Campus Example: Think of a concert venue. The security checkpoint (metal detectors) is the drum — it's the slowest step and sets the pace of entry. The buffer is the roped-off queue area in front of security that keeps a steady stream of people ready. The rope is the staff at the outer gate who control how many people they let walk toward security — if the queue buffer is full, they slow the inflow.

Without DBR, upstream stations produce as fast as they can, WIP piles up before the bottleneck, and chaos follows.

12 What are the 5 TOC Focusing Steps?

1. IDENTIFY the system's constraint. 2. EXPLOIT it — squeeze every drop of capacity from it (no idle time, no bad parts). 3. SUBORDINATE everything else to that decision — non-bottlenecks serve the bottleneck's pace. 4. ELEVATE the constraint — invest to increase its capacity. 5. REPEAT — if the constraint has shifted, go back to step 1. Don't let inertia become the new constraint.

🏫 Campus Example — Course Registration:
1. Identify: The server handling registration requests crashes at 7 AM every semester — it's the constraint.
2. Exploit: Spread registration windows by class year so the server never hits 100% load simultaneously.
3. Subordinate: Advising appointments are scheduled to finish before each group's registration window.
4. Elevate: Invest in more server capacity or a cloud-based system.
5. Repeat: Now the constraint might shift to advisor availability — start the cycle again.

13 What's the difference between a process batch and a transfer batch — and why does it matter?

A process batch is the total number of units produced in one setup. A transfer batch is the portion of the process batch that moves to the next station without waiting for the whole batch to finish.

🍕 Campus Example: A pizza station makes 200 pizzas in one shift (process batch = 200). But you don't wait until all 200 are done to start serving. You transfer them in groups of 20 (transfer batch = 20). This keeps downstream serving stations busy and customers fed — rather than everyone waiting 3 hours for the full batch.

Making process batches too large is a common cause of moving bottlenecks — a non-bottleneck machine hogs time on a huge batch, starving downstream stations and temporarily acting like a bottleneck.

14 How do JIT, MRP, and Synchronous Manufacturing compare?

Dimension	JIT	MRP	Synchronous Mfg.
Best environment	Continuous flow, make-to-stock	Job shop, custom shop	Job shop, custom shop
WIP levels	Very low	Very high	Low
Cycle time	Very short	Very long	Short
Schedule flexibility	Level production for 30+ days	Frozen ~30 days, variable in workcenters	Can change daily
Capacity awareness	Tries to balance	Starts OK, quickly inaccurate	Founded on capacity limits
Quality approach	Every station responsible	Inspect where cost justifies	Inspect before bottleneck + after
Scheduling direction	Pull (Kanban)	Backward (from due date)	Forward (from drum)

🎯 Key Takeaway: Synchronous manufacturing is the most flexible of the three — it can adjust daily, it's capacity-aware by design, and it focuses quality control where it matters most: protecting the bottleneck from defective inputs.

✓ Check Your Understanding

5 quick questions covering both chapters. Click an answer to check yourself.

Q1 Which priority rule minimizes average flow time for a single machine?

AFCFS — First Come, First Served

BSOT — Shortest Operating Time

CEDD — Earliest Due Date

DCR — Critical Ratio

Q2 In the Drum-Buffer-Rope system, what does the "rope" represent?

ASafety stock placed before the bottleneck

BThe bottleneck resource itself

CUpstream communication to control material release

DThe final quality inspection station

Q3 A non-bottleneck machine is scheduled with oversized batches. What's the most likely consequence?

AIt starves downstream stations and acts like a moving bottleneck

BThroughput increases because setups are reduced

CWIP decreases due to fewer changeovers

DIt has no impact since it has excess capacity

Q4 Why is service scheduling (like staffing a campus café) fundamentally harder than manufacturing scheduling?

AServices have more machines to schedule

BService workers are less skilled

CManufacturing has more variable demand

DYou can't control when customers arrive or smooth demand easily

Q5 According to TOC, what should you do FIRST when you discover a system constraint?

ABuy more equipment to add capacity

BExploit it — maximize its utilization with no wasted time

CBalance all other resources to match its capacity

DMove it to a different location in the process

0 / 5

Questions Correct