Steel Mill Production Software: Custom Build Versus Off the Shelf

What the off-the-shelf products actually do well

Coil 44821 is the whole argument, so start with who already solves part of it.

PSI Metals is the reference planning and production suite in this market and it deserves that position. It genuinely models order to material matching, applied inventory pools and re-application, which is the thing every generic system misses. If your mill has the appetite for a long implementation and a configuration model rather than a code model, PSI is the serious answer and we say that to mills regularly.

SAP, usually S/4HANA, is excellent at purchasing, finance, sales order administration and the batch level lot tracking that satisfies most auditors. It is blind between the caster and the shipping bay, which is not a defect, it is scope. Nobody should rebuild SAP.

Your level 2 systems from Primetals Technologies, SMS group and Danieli Automation are better at their own lines than anything you would write. They hold setup models, pass schedules, thermal history and process data at a resolution nobody else can match, and each supplier offers a layer above its own equipment that works well as far as that equipment reaches. AVEVA and Siemens sit in the manufacturing execution layer with an ISA-95, that is IEC 62264, model that is a reasonable starting shape for a plant.

If your operation is a single line re-roller buying hot band and shipping to stock, stop reading. An enterprise resource planning (ERP) system with lot tracking, one good spreadsheet and a disciplined planner will hold you for years, and a build is capital better spent on a leveller or a slitter.

Where they stop: material that changes identity on the way through

Steel is a divergent process with continuous re-identification. One heat becomes several slabs. One slab becomes a coil. One coil becomes child coils and then cut lengths. At every step the material can change grade, owner, destination and value. Standard manufacturing software models convergence and treats material as fungible against a part number. In your mill nothing is fungible, and that single mismatch is where generic products come apart.

The concrete case: a coil comes off the temper mill with an edge defect over the last eight tonnes. Quality downgrades it from an exposed automotive grade to commercial quality. It was applied to an order closing Thursday. Four questions now need answering inside an hour. Is there another coil in the applied pool that meets the specification. If not, does a slab exist that can be rolled in time. Can the downgraded coil serve another open order inside tolerance, or does it join secondary inventory ageing in the yard. And what do you tell the customer.

SAP will allocate a batch to a sales order. It will not tell you that the downgraded coil satisfies four other open lines inside thickness and width tolerance, one of which is late, and that applying it there beats a discount in secondary.

The second thing that stops them is genealogy across splits. Your customers want a mill test certificate under EN 10204 3.1 tying actual chemistry and mechanical results to the specific piece they received, and automotive and pressure vessel customers audit that chain without gentleness. A coil slit into four inherits from one heat, which is easy. A coil rolled from a transition slab in a mixed grade cast sequence inherits from two, which is not. Level 2 systems track the piece on their own line, and the joins between lines get reconstructed later from timestamps by somebody with database access and a free afternoon.

The arithmetic: named seats against a one-time build

Metals planning suites are licensed by module and by named user, with implementation quoted separately and measured in quarters. Put your own quote into this rather than ours.

Suppose the figure is $300 a named user a month. Forty seats across planning, scheduling, quality, metallurgy and shipping is $144,000 a year. Metals implementations are long enough that five years is the honest horizon, so that is $720,000 in licence alone, before implementation, before the annual maintenance percentage, and before the change requests that follow every practice change.

A first release build at $180,000 with $30,000 a year of support is $300,000 across the same five years. The crossover sits near 17 named seats at $300 a month, and near 34 seats if your quote lands closer to $150. Almost every integrated mill is well past that.

Transactions matter more than seats in one place. If a vendor prices by tonne shipped or by piece tracked, work out the fee at your annual tonnage and then at the tonnage in your five year plan. A fee that scales with output scales with the thing the mill exists to increase.

The number that does not appear on either side of that comparison is the one that usually decides it. Material sitting in secondary inventory for six months is capital you already spent, and continuous re-application is what releases it. One mill's answer to how many tonnes are ageing in the yard right now is a better business case than any licence comparison.

What a custom build actually costs

A first release runs $120,000 to $250,000 over 16 to 24 weeks. That covers the plant wide piece and event model, level 2 acquisition for the lines that matter most, order to material matching with continuous re-application, and genealogy with certificate generation.

A full platform adding caster and rolling campaign scheduling, the remaining lines, yield costing, secondary inventory and customer portal access runs $350,000 to $900,000 phased over 12 to 24 months.

Data migration is 10 to 25 percent of the build figure, and steel sits high in that band. You are not moving rows, you are reconstructing a genealogy graph from three level 2 databases whose piece identifiers disagree, and identifier reconciliation is the line item that surprises every mill. Year two costs 15 to 20 percent of the build annually, covering practice rule changes, new customer specifications, a level 2 upgrade you did not schedule, and the certificate format a new automotive customer insists on.

What drives the number up specifically here: the count of production lines and how many different level 2 suppliers are represented, because each acquisition integration is separate work. Whether piece identifiers are consistent across lines or need reconciling. Whether chemistry and mechanical test results arrive from a laboratory information system or from a spreadsheet. And whether customer specific tolerance envelopes exist as data anywhere, or only in a metallurgist's head.

The four situations where building wins

Regulatory and certification fit. If you certify under EN 10204 3.1 and 3.2, ship into automotive under IATF 16949 with production part approval evidence, or supply pressure vessel grades where the chain from heat to piece is audited, a genealogy graph you own is a compliance asset rather than a convenience. Append only event logs matter here, because a genealogy record that can be silently edited is worth nothing in an audit.

Scale economics. Past roughly 17 named seats on a per-user product, or past 300 open orders at a time, the licence arithmetic has already turned and the operational case is separate and larger.

A workflow that is your advantage. Practice rules are yours alone: which grades follow which on the caster without a tundish change, how wide a width jump the hot strip mill accepts in a campaign, which coating line campaigns run weekly. Hundreds of these rules exist as the knowledge of two or three schedulers. Encoding them as data that schedulers edit without a release is the feature operations directors underestimate at kickoff and value most a year later, because when a scheduler retires the mill keeps its sequencing logic.

Integration sprawl across three or more systems. Caster from one supplier, hot strip mill from another, galvanising line from a third means three data models, three timestamp conventions and three opinions about what a production event is. An acquisition layer per line normalising into one plant wide piece and event model is where a large share of the effort goes, and it is what makes everything else possible.

How to decide in a week

Do not run a vendor demonstration. Run this instead, using last month's real data.

Take every downgrade event from the last thirty days. For each one, reconstruct on paper what the best available re-application was at the moment quality made the call, and compare it to what actually happened. Then price the difference: tonnes that went to secondary at a discount when an open order inside tolerance existed, plus tonnes still ageing in the yard.

• If the gap over thirty days is under the cost of two planner seats, buy the package and configure it properly.
• If the reconstruction itself takes more than a day because the data lives in three level 2 systems, your first project is the acquisition layer, not the planner.
• If you cannot produce a certificate for a coil rolled from a transition slab without a metallurgist rebuilding it, build the genealogy graph first.
• If your schedulers cannot write down the practice rules in an afternoon, that knowledge is a single point of failure and it belongs in a system.

Then buy a paid discovery phase rather than accepting a proposal. At Digital Heroes that is two to three weeks producing a signed product requirements document covering the piece and event model, the level 2 interfaces line by line, the certificate templates and the acceptance criteria. You own that document either way, and you can tender it against three firms.

Where we are the wrong choice: if you want a configured metals suite with a vendor consultant on site for a year, PSI Metals is the better fit and we will say so on the call. We are a team of more than fifty specialists with over 2,000 projects delivered, verifiable on Clutch, Trustpilot, Fiverr Vetted Pro and by D-U-N-S number. Our own products, ShopScore, HeroCheckout and Section Vault, are commerce software, so judge us on delivered client work rather than on a metals product we do not sell. We hold India LLP, US LLC and UK LTD entities, so the intellectual property assigns under your own law. You meet the engineers who will do the work before you sign anything.

Book a 30-minute call with Digital Heroes and get a written plan and a fixed quote within 48 hours.