Plastics Scheduling: Modeling Mold Cavities and Pieces-Based Capacity

Plastics cavity capacity scheduling models a multi-cavity mold or a bank of presses as machine instances and schedules the work by hours or pieces per hour, so EDGEBIC by User Solutions plans against real molding throughput instead of a single-cavity time. For an injection molding shop, the whole problem is that a four-cavity mold is not four separate jobs and not one slow job either. This post is about modeling cavities and pieces correctly, so the plan matches what comes off the press.

The general question of loading work against real, limited capacity is finite versus infinite capacity scheduling, and the plain definition of a schedule is what is production scheduling. Here we translate both into the units a molding floor lives in: cavities, presses, and pieces per hour.

The cavity that breaks a naive schedule

A routing that says a molding step runs eight hours is fine for a single-cavity tool. Put a four-cavity mold on it and the number is wrong four ways: the cycle produces four parts per shot, so the throughput is four times higher, but the routing still charges the single-cavity time. A plan built on that either massively overstates the time a run takes, and idles the press, or understates it and blows the date.

Plastics shops also measure in the wrong currency for most schedulers. The floor thinks in parts per hour and shots per shift, not machine-hours. A plan that can only reason in hours forces the planner to convert everything by hand, and the conversion drifts the moment a cycle rate changes or a mold moves. The right model lets the schedule reason in cavities and pieces directly.

Modeling cavities as capacity, not as steps

The clean way to represent parallel cavities or a bank of identical presses is as instances on the work center. An instance is one unit of physical capacity, and a work center's instance count tells the scheduler how many units it can load at once. This is deliberately not the same thing as a machine pool of different-but-interchangeable machines, which is covered in plastics machine pools; instances are identical units under one calendar and one speed, which is exactly what a multi-cavity mold or a set of matched presses is.

With load balancing on, the default, the scheduler divides an operation's hours evenly across the instances so they finish together instead of stacking on one. A 20-hour block on a four-instance work center becomes five hours on each instance, all starting together and done in a quarter of the single-unit wall-clock time.

The plan now reflects the throughput the cavities actually deliver, without splitting the routing into four fake operations. When a cavity count is not a whole number, for example a half-capacity backup unit, the scheduler supports fractional instances and redistributes the partial unit's share across the full ones, so a two-and-a-half-cavity setup is modeled honestly rather than rounded.

Scheduling in pieces per hour

When the real constraint is throughput rather than time, express the step in pieces. A routing step can carry a pieces-required figure and a pieces-per-unit multiplier, so the effective piece count scales with order quantity. The work center carries the capacity type that says it is measured in output, loaded through the work center import and shown as a read-only column on the grid, and the rate itself lives on the step. Put the cycle the press really sustains there and the plan reasons in molded parts rather than in a nominal cycle.

For a molding floor this closes the gap between the routing sheet and the shop language. You state how many parts an order needs and how fast the press makes them, and the schedule works in those units end to end. It is the same piece-based thinking that drives lot streaming, where downstream operations start after a transfer batch of parts is ready rather than after the whole run finishes, described in what a transfer batch is.

Headroom: the lever for a hot press

Presses rarely sustain nameplate cycle for a full shift. Cooling, minor stoppages, and the difference between the tool's rated cycle and its real one all eat into throughput. That gap has to live somewhere in the model, and it is not a percentage on the work center: work centers run at 100 percent utilization and the value is not editable on the work center screen.

So put it in the calendar, where it is visible. A press whose eight-hour shift really yields six and a half productive hours should carry a shift that says so, and with two instances it then offers 13 hours of capacity instead of 16. For losses that recur on a known pattern, a downtime event takes the hours off automatically. Capacity is built at the start of a scheduling run, so make the change before you schedule; editing master data never moves bars already placed.

For a specific day you know will run short, a scheduled color or mold change that eats the morning, a per-day capacity override sets an exact figure for that one date and replaces the formula outright. It also stores a reason, so the next planner reading the plan can see why the day was short.

Presses that run one job per day

Some molding cells dedicate a press to one job for the whole shift because the mold or color change occupies the day. The one-per-day setting handles that. With more than one instance, each new job is assigned its own instance rather than pooling the capacity, so three six-hour jobs land on three separate presses at the start of the shift rather than queuing on one. This is the same instance-selection logic that runs a furnace shop, applied to a press bank where changeover discipline matters. It composes naturally with mold changeover scheduling, which decides how long each change takes, while the one-per-day rule decides how the day's jobs are distributed across the presses.

Putting cavities, pieces, and headroom together

A molding floor modeled well looks like this: cavities and matched presses as instances so throughput is real, pieces and a sustained cycle on the step where the floor thinks in parts, shifts that describe what each press actually delivers, and one-per-day where a press is dedicated for the shift. Load the shifts each press runs and the plan spreads work across cavities, promises dates the floor can hit, and moves rush orders through the real capacity picture rather than a nameplate one. It sits alongside the multi-shift model that molding shops use to run around the clock.

This is the capacity realism User Solutions has brought to manufacturers since 1991, the difference between a plan that assumes best-case cycles and one that matches the press. A schedule that speaks cavities and pieces is a schedule a molding floor will actually trust.

See it model your worst multi-cavity tool. Bring a molding routing and a press to a demo and watch the plan schedule the cavities as real capacity. The EDGEBIC by industry guide shows where this fits alongside the rest of a plastics operation.