The Constraints That Actually Bind
On paper capacity is machine hours. In practice the binding constraints are die availability, alloy furnace assignment, changeover time and downstream trim and secondary capacity. A die can only be on one machine, and a die in repair is unavailable regardless of open machine time.
Alloy matters because switching a cell from A380 to a different alloy means cleanout and lost production. Trim presses and machining often run slower than casting, creating a queue after the casting step that does not show on a machine-only schedule. List every constraint before choosing a tool.
- Machine tonnage and shot size compatibility per die
- Die status: running, in repair, awaiting insert, in storage
- Alloy and furnace assignment constraints
- Trim, machining and secondary capacity after casting
Changeovers and Sequencing
Die changes consume machine time and have warm-up scrap. Grouping jobs by alloy and by die family reduces changeovers, but too much batching builds inventory and risks late shipments. A good schedule balances lot sizing against customer due dates, using realistic changeover times by die and machine, not a single flat figure.
Where customers pull on a kanban or releases against blanket orders, sequence around the release dates and build a small buffer for the most variable parts. Record actual changeover times each time to keep the plan honest.
Where ERP Scheduling Falls Short
Standard MRP assumes infinite capacity and fixed lead times. Systems such as Infor SyteLine, Epicor Kinetic, Plex and Global Shop Solutions offer finite scheduling or APS modules, but they only work if the data behind them is right: accurate cycle times, die-to-machine eligibility, and reliable die status. Many die casters therefore run a hybrid, with ERP for orders and a spreadsheet or whiteboard for the actual cell schedule.
That hybrid is fine, if the spreadsheet is fed from ERP and updated actuals flow back. Otherwise, the two diverge within days.
- Maintain die-to-machine eligibility as master data
- Link die maintenance status to schedule availability
- Feed actual cycle times back to the planning data
- Show scheduled versus actual by machine daily
Try it here, free
Barrel Residence Time Calculator
Estimate how long resin sits in the barrel and how much of the machine capacity a shot uses, to spot degradation risk from oversized presses.
What AI Can and Cannot Do Here
AI can help in a narrow, practical way: answering questions such as which jobs are at risk if Die 214 stays down two more days, which open orders could run on the 600 ton machine instead, or what the planned loading looks like next week by machine. It can read ERP orders, die status and cycle time history and propose a revised sequence for the scheduler to approve. It should not auto-publish a schedule or move work without human review.
Optimization engines can also help, but they depend on the same clean master data as any other scheduler.
When Finite Scheduling Is Overkill
If you run a few dedicated dies on dedicated machines with steady customer releases, a simple weekly board is enough and a finite scheduling project will cost more than it saves. Consider it when die count is high, machines are shared across many dies, expedites are frequent, or late shipments trace back to planning surprises. Before buying anything, track for a month why jobs shipped late.
If most causes are die repair and quality holds rather than sequencing, fix those first.
Frequently asked questions
How do you schedule die casting machines?
Assign each job to a machine that fits tonnage and shot size, confirm the die is available and serviceable, group by alloy and die family to limit changeovers, and check trim and secondary capacity downstream. Then compare to due dates and adjust. Finite scheduling tools automate this if the master data is accurate.
How long does a die change take?
It depends on die size, machine, quick-change systems and crew, ranging from under an hour to several hours. Record your actual times by die and machine rather than assuming a flat number, since this drives lot size decisions.
Should die status be in the ERP?
Yes, ideally. A scheduler should not plan a job on a die that is awaiting repair. At minimum, keep a die master with status, location and shot count, and update it when maintenance starts and finishes.
Is APS worth it for a small die caster?
Often not at first. Clean master data and a disciplined weekly schedule deliver most of the benefit. Look at APS when die and machine combinations multiply and manual sequencing cannot keep up.
If your schedule lives in a spreadsheet beside the ERP, we can show how a private assistant can read both and answer what-if questions about die and machine loading.