In this guide 6 sections
Buy ready software when your plant runs standard jobs and the package already records shot counts, die life, machine downtime and rejects by cause. Build custom when die maintenance rules, customer part marking or machine data feeds do not fit any package you trial, however long you tune its settings. Test both on one real week.
Start from the records, not the feature list
Three records carry a die casting plant. Each die has a shot count, each machine has a run log, and each rejected casting carries a cause, while everything else is support work around those three.
Write those records down first. List the fields on your present shift sheet, the die history card and the rejection tag, because sales brochures cannot tell you whether a package holds them.
Mark each field as must-have or nice-to-have, since a package that misses one must-have field will push the supervisor back to paper within a month.
- Die number, cavity count, shots since polish
- Machine, shift, operator and cycle time per part
- Reject causes: porosity, cold shut, flash, short fill
- Melt temperature and alloy lot per shift
Where ready software usually fits
Ready software suits the common spine of a plant, because orders, ingot and finished casting stock, dispatch papers, alloy purchase and inward quality checks follow similar patterns everywhere.
If your shop works from customer drawings and ships in returnable bins, a foundry package will likely cover most of it, including piece rate invoicing. Setup is quick and training is light.
Request a trial with your part list, then watch how it treats one part cast on two dies, since many packages assume one die per part.
Where custom starts to earn its cost
Custom pays off when rules are odd. A die may go for maintenance at a shot limit that varies by part, or a customer may want every casting marked with a heat code and die cavity.
Machine connections are another case. Reading shot counters straight from a cold chamber machine controller saves the clerk from copying numbers, but each controller brand speaks differently, so a package rarely covers the mix.
Custom also suits a plant that lends dies to a job work partner, since tracking where a die sits and what shots it ran there needs its own screens.
- Die maintenance triggers by part and alloy
- Cavity marking and traceback on complaints
- Controller feeds from old and new machines
- Die loans to outside toolrooms and polish shops
Compare the two on one real week
Pick a week with a die change, a breakdown and at least one customer rejection, then replay it through each option, because clean demo data proves little.
Score each option on four things. Time to enter a shift, whether the die shot count stays right after a swap, whether reject causes roll into a weekly Pareto, and what the manager sees unasked.
Count the people cost as well. A custom build needs an owner on your side who sets priorities, while a package needs someone who learns its settings and keeps in touch with the vendor.
Decide in stages and keep the exit open
You rarely choose all or nothing. Many plants buy a package for orders and stores, then add a small custom module for die life and reject analysis once the shop floor trusts the first records.
Insist on data export in plain files such as CSV, so that pulling your shot counts and reject history out later costs far less, and test the export early.
Your next step is simple. Print the die history card, the shift sheet and the rejection tag, circle every field, and hand that stack to each vendor you speak with.
Build or buy scorecard for a die casting plant
One sheet, three columns. Record, ready package and custom build. Rows cover shot count, cavity marking, run log, reject causes, alloy lot, maintenance trigger and die loans, each cell scored from zero to three and totalled at the foot.
Open a blank worksheet to printQuestions owners ask
Can one package handle both die casting and machining?
Some can, when routings allow several operations after casting. Check that a casting lot keeps its identity through trimming, shot blasting and machining, so a reject at the last step still points back to a die.
Do we need machine connections from day one?
No. Operators can key shot counts at shift end. Add controller feeds later, only for the machines where manual entry keeps causing arguments, missed shots or gaps in the die record.
How long does a custom build take for a small plant?
It depends on scope. A first release covering only die records and reject causes can stay small, while machine feeds and die loans arrive in later stages. Ask for a staged plan, so supervisors use the first release early and shape the later ones from what they find missing.
What if our dies are owned by the customer?
Then each die needs an owner field and a tooling register tied to the customer, so you can see which customer’s dies are near their shot limit. Test the die life report per owner.