Onboarding a New Work Center and Rebalancing the Plant in EDGEBIC

Onboarding a new machine in EDGEBIC by User Solutions is four steps: create the work center, give it a shift so it has capacity, add it to the routing steps that should use it, and reschedule, at which point the overloaded machine's load drops and the new one picks up its share. This walkthrough adds a second mill to a plant whose single mill is overbooked and watches the rebalance happen on the report. It is part of our worked examples series, and it starts where the month-end capacity review ends: a machine reading CRITICAL that you have decided to relieve.

The Setup: One Overbooked Mill

Summit Gearworks runs five work centers on a day shift. The capacity review flagged the single mill, MILL-1, at CRITICAL: 128 scheduled hours against 100 available, 28 hours over. Management approved a second mill, MILL-2, physically identical, same speed. The plan is to bring it online and split the milling load. The numbers are illustrative; the onboarding steps and the rebalance behavior are the documented behavior.

Step 1: Create the Work Center

Priya opens Work Centers and clicks to create a new one. She enters the name (MILL-2), a description, and sets the physical properties to match MILL-1: capacity type hours, efficiency 100%, and Number Of Instances matching the number of physical spindles (one, in this case). She saves.

At this moment MILL-2 exists in the plant list but does nothing. It has no shift, so it has zero available hours, and no routing names it, so no job can reach it. Both gaps have to close before it earns its keep.

Step 2: Assign a Shift So the Machine Has Capacity

Priya opens MILL-2's shift assignment. MILL-1 runs the plant-wide day shift, so she assigns the same global day shift to MILL-2. Now the machine has real capacity: the same 8 working hours a day the rest of the plant runs.

If MILL-2 ran different hours, she would instead configure work-center-specific shift hours to override the global shift, for example a ten-hour day or a Saturday. And because holidays and downtime are per machine, a future maintenance day on MILL-2 will zero only its capacity, leaving MILL-1 and everything else untouched. Assigning the shift is the step people skip, and a machine with no shift silently reads NO-CAPACITY on the utilization report and never receives a single hour.

Step 3: Add the New Mill to the Routing

Capacity alone does not move work. The engine only routes a job to a machine the routing names. Priya opens the milling step on the affected products' bills of routing and decides how MILL-2 should be used. She has two documented options:

• Independent parallel. Set the milling step's work center type to independent parallel and add MILL-2 as an alternative with category Parallel and factor 1.0 (same speed). The engine then finds a window where both mills have capacity and splits the step's hours across them using load balancing, so the job finishes in roughly half the wall-clock time. This is the right choice when she wants both mills working the same jobs.
• True alternative. Add MILL-2 as a true alternative instead, so the engine routes a milling job to MILL-2 only when MILL-1 is fully booked. This is the right choice when MILL-1 is preferred (say it is newer) and MILL-2 is overflow capacity.

For a straight rebalance of an overbooked machine, Priya chooses independent parallel on the highest-volume products, so the load genuinely divides. She saves the routing.

Step 4: Reschedule and Rebalance

Priya sets the scheduling mode to one that reschedules existing work (a FullReschedule gives the cleanest rebalance here, since the goal is a clean-sheet redistribution across the new capacity) and clicks Schedule / Re-Schedule. The engine now sees two mills on the milling steps. For each parallel milling operation it finds a shift window where both MILL-1 and MILL-2 have capacity, splits the hours across them, and emits two schedule lines per step, one per machine.

The 128 hours that used to pile onto MILL-1 now spread across two machines. Roughly 64 hours each, and the milling jobs finish sooner because two machines run them at once.

Where the Load Lands

Priya confirms it two ways. On the Job View, she filters to a milling job and expands it: two schedule lines for the milling step, one on MILL-1 and one on MILL-2, each carrying about half the step's hours, overlapping in time. On the Work Center Utilization report, MILL-1's rating has dropped from CRITICAL to GOOD and MILL-2 now appears in the table at a matching load. The plant's overloaded-machine count in the KPI strip drops by one.

What This Onboarding Proves

1. A new machine needs three things before it works: existence, a shift, and a place in the routing. Miss the shift and it reads NO-CAPACITY; miss the routing and it stays idle.
2. Capacity does not route itself. The engine only sends work to a machine the BOR step names, as parallel (split) or true alternative (overflow).
3. The rebalance is measurable. Utilization on the overloaded mill dropped from CRITICAL to GOOD, and the report shows both mills carrying the load.
4. Parallel splits the hours; alternative catches overflow. The routing choice decides whether both machines share every job or one backs up the other.
5. Per-machine calendars keep it clean. MILL-2's future maintenance day will zero only its own capacity.

Variations Worth Trying

Overflow instead of split. If MILL-1 is genuinely preferred, configure MILL-2 as a true alternative and reschedule. Milling jobs run on MILL-1 until it fills, then spill to MILL-2, which keeps MILL-1's setups stable while still relieving the overload. That fallback behavior has its own backup-machine walkthrough.

Different speed. If MILL-2 is slower, set its parallel factor above 1.0 so the engine costs its hours honestly, or its per-alternative setup and runtime if it is a true alternative. The split then reflects the real speed difference rather than assuming the machines are twins.

Group the mills as a pool. If you have several interchangeable mills, a work center group lets a routing step target the pool and the engine picks the member by strategy, which scales better than naming each machine on every step. See what a work center group is.

The Bigger Point: Expansion Is a Data Change, Not a Rebuild

Adding capacity to a spreadsheet-run shop means re-drawing every affected plan by hand. In a finite capacity engine it is a master-data change plus one reschedule: create the machine, give it hours, name it in the routing, and let the engine redistribute the load. The rebalance you were hoping for shows up as a number on the report, so you can prove the new machine paid for itself.

User Solutions has built that discipline into scheduling tools since 1991, for operations that add and retire capacity constantly: the US Navy, GE, BAE Systems, and Cummins among them. GE Railcar's move from 30% to 90% on-time delivery came from exactly this kind of load rebalancing against finite capacity. EDGEBIC carries that lineage forward.

Bring your overbooked machine and the one you are thinking of adding. Contact us and we will model the rebalance against your routing data, or read the splitting-across-parallel-machines walkthrough to see the hour-split mechanics in detail.