Comparing Two What-If Scenarios for a Capacity Decision in EDGEBIC

Comparing two capacity options in EDGEBIC by User Solutions is a controlled experiment: model each change to the master data, reschedule the same jobs, and read the same three numbers (late jobs, constraint utilization, at-risk completion dates), so the decision is a measured comparison instead of an argument. This walkthrough runs add-a-shift against add-a-machine for an overloaded mill and picks a winner with numbers. It is part of our worked examples series, and it turns the read-only findings of the month-end capacity review into a spend-or-not decision.

The Setup: An Overloaded Mill and Two Ways to Fix It

Granite State Machine has a mill (MILL-1) reading CRITICAL: too many jobs, not enough hours, and six orders projected late this month. Management will either run a night shift on the mill or buy a second mill. Both cost money; only one gets funded. The operations lead, Ken, will decide with the engine. The numbers are illustrative, but the compare-two-reschedules method is exactly how you would run it.

Step 0: Capture the Baseline

Before changing anything, Ken records where things stand, because a comparison needs a starting point. He runs two reports on the current schedule:

• Late Jobs report: 6 orders late, ranked by lateness, with a blocker hint on each.
• Work Center Utilization report: MILL-1 at CRITICAL (roughly 128%), everything else GOOD or LOW.

These two numbers, late count and mill rating, plus the specific completion dates of the most at-risk orders, are the scoreboard. Every option is judged on the same three.

Step 1: Model Option A, Add a Night Shift

Ken creates a night shift and assigns it to MILL-1, so the mill now runs day and night. This roughly doubles the mill's available hours without adding a machine. He sets the scheduling mode to reschedule existing work and runs it. The engine replans the affected jobs against the new, larger mill capacity.

Then he reads the scoreboard again:

• Late Jobs: down from 6 to 0.
• MILL-1 utilization: dropped from CRITICAL to HIGH (say 88%).
• At-risk orders: all now complete on or before their due dates.

Option A clears the lateness. The mill is still busy (HIGH), which means the night shift is well used but leaves little headroom for growth. Ken records these numbers and reverts the change so the next option starts from the same baseline.

Step 2: Model Option B, Add a Second Machine

Ken now models the alternative: create a second mill (MILL-2), assign it the day shift, and add it to the milling routing steps as an independent parallel or true-alternative work center so jobs can actually reach it. He reschedules the same jobs.

The scoreboard:

• Late Jobs: down from 6 to 0.
• MILL-1 utilization: dropped from CRITICAL to GOOD (say 64%), with MILL-2 also at GOOD.
• At-risk orders: all complete comfortably ahead of their due dates.

Option B also clears the lateness, and it leaves real headroom: two mills at 64% each can absorb more work, and milling jobs finish sooner because two machines run them at once. The onboarding steps for that second machine have their own work-center rebalancing walkthrough.

The Comparison

Both options solve today's problem. The decision now turns on the trade the numbers expose: Option A fixes lateness at recurring labor cost with little slack, while Option B fixes it at capital cost with room to grow. If the overload is a temporary spike, the night shift is the leaner choice. If demand is trending up, the second mill buys future capacity the shift cannot. Ken is choosing between two known outcomes, not gambling on one.

Step 3: Watch for a Moved Bottleneck

One check keeps the decision honest. Relieving the mill can expose the next-tightest machine. After each option, Ken confirms on the utilization report that no other machine turned CRITICAL. If, say, the grinder jumps to HIGH once the mill stops being the constraint, he has found the new bottleneck, and the real question becomes whether either mill fix matters until the grinder is addressed. The reports show exactly where the constraint went, so the comparison never hides a moved problem.

Step 4: Commit the Chosen Option

Once Ken picks, say Option B, he leaves that master-data change in place (second mill created, assigned, and routed) and runs a final reschedule so the live plan reflects the decision. The what-if is now the plan. He documents the baseline-versus-chosen numbers as the justification for the spend.

Because a capacity what-if reschedules the live plan when it runs, Ken treated the whole exercise as deliberate: baseline captured, each option run and recorded, then the chosen final state committed. For a purely read-only estimate that writes nothing, the quote simulation is the alternative tool, though for a plant-wide capacity decision the full reschedule is what gives the true late-jobs count.

What This Comparison Proves

1. A capacity decision is a controlled experiment. Same engine, same jobs, same due dates, one variable changed at a time.
2. Three numbers decide it. Late-jobs count, constraint utilization, and at-risk completion dates.
3. Both options can solve the problem differently. The night shift clears lateness with little slack; the machine clears it with headroom.
4. The reports catch a moved bottleneck. If a different machine turns red, the comparison shows it before you spend.
5. The winner is a table, not a hunch. You commit the option whose numbers you can defend.

Variations Worth Trying

Add capacity to only the worst days. Instead of a full night shift, override the mill's daily capacity on the specific red days from the heatmap and reschedule. If a handful of overtime days clears the lateness, that is cheaper than either standing option.

Throttle low-priority work. Schedule the dateless stock jobs at reduced utilization so they stop crowding the mill, reschedule, and see whether the overload was really demand or just sequencing.

Test a demand increase. Add a hypothetical big order to each option and reschedule, so you compare not just today's fix but how each option handles growth. This is where the second mill's headroom shows its value.

The Bigger Point: Decide Capacity by Rehearsal

Capacity decisions go wrong when they are made from a spreadsheet that cannot reschedule. A finite capacity engine lets you rehearse each option against the real order book and read the outcome, so you are choosing between measured futures instead of defending opinions in a meeting. The method is simple and repeatable: capture the baseline, change one thing, reschedule, read the same numbers.

User Solutions has built that discipline into scheduling tools since 1991, for operations where capacity decisions are constant and expensive: the US Navy, GE, BAE Systems, and Cummins among them, the last across 33 locations. EDGEBIC carries that lineage forward, and the results guide frames these comparisons as a business case.

Bring your overloaded machine and the two options on the table. Contact us and we will rehearse both against your real order book, or read the quoting-three-routing-options walkthrough for the same simulate-and-compare discipline applied to a customer promise.