Scheduler vs ERP vs Configurable APS for Process Plants

Introduction

If you've got a scheduler on the floor, planning probably feels handled. The Gantt is clean, the line is sequenced, the week looks under control. It isn't handled, and the reason is worth understanding, because the scheduler is only one of three things a process plant plans with, and none of the three covers the whole problem.

Almost every process manufacturer has the same base: an ERP, with the real planning happening in spreadsheets. Some add a scheduler on top. A few buy a configurable Advanced Planning and Scheduling (APS) suite. Each layer solves a piece, and each leaves a different piece unsolved. Here is the honest version of what each does well, where each stops short, and where a custom-built model fits the part none of them reach.

ERP plus spreadsheets: arithmetic, not optimization

This is the base layer almost every process plant has. The ERP holds the orders, the inventory, and the bills of material, and it nets requirements: demand minus stock, inflated by a yield factor, exploded through the BOM. That's arithmetic. It tells you what's needed, not the best way to make it. So the actual plan, the sequence, the blend, the changeover order, the rate, gets built in a spreadsheet by a planner who knows the plant. It works until that planner is out, or the SKU count grows, or the month gets complicated. The ERP never saw the physics, and the spreadsheet can only solve one piece at a time. There's no optimization anywhere in the stack, just a system of record plus a very capable person holding the rest together by hand.

An ERP nets requirements. It doesn't optimize a plan. The optimization is happening in a spreadsheet, or it isn't happening at all.

A scheduler: a sharp schedule for a plan nobody optimized

The common next step is a dedicated scheduler or an MES scheduling module, sitting on top of the ERP. It's a real improvement on the floor. It sequences a line, respects a changeover matrix, and produces a clean Gantt. If the daily sequence on one set of equipment is your only pain, it earns its keep, which is exactly why a plant with one can feel finished. But a scheduler optimizes the floor in isolation. It takes the demand it's handed and orders the jobs, without seeing purchasing, inter-plant transfers, multi-site capacity, or the distribution and shipment side. It never asks whether that demand mix was the right thing to build, whether the raw material is there to feed it, or whether a different plant should have made it. You end up with a sharp schedule for a plan nobody optimized. The sequence is tight; the plan behind it is still a spreadsheet.

A scheduler answers "what order do I run these jobs." It never asks whether those were the right jobs, fed by the right purchases, made at the right site. That's the plan, and the scheduler doesn't own it.

Configurable APS: end to end, but built on a template

At the top sits the configurable APS or Integrated Business Planning (IBP) suite. These do plan end to end, and for a large enterprise with the budget and the team, they work. This isn't a knock on the category. An APS that plans the whole chain is the right idea, and WonForge is one too. The problem for a mid-market process plant traces to one root: these platforms are configured from a template, not built for you. The template was designed for discrete manufacturing, so the process physics, the sequence-dependent changeovers, the variable yields, the blends, the co-products, gets approximated into the template or pushed back out into a spreadsheet. And the configuration is exactly what makes them expensive: the six- or seven-figure license plus a configuration project that runs many months to years. It's the cost of bending a generic template toward your plant, and it's the reason most mid-market process manufacturers never bought one. You can spend a fortune and still reconcile the physics by hand, because the platform was never built for how a process plant actually runs.

A configurable APS plans end to end, but it's configured from a discrete template, so the process physics still leaks into spreadsheets, and the configuration itself is what makes it cost six figures and take many months to years.

Where WonForge sits

WonForge is an APS too. The difference is that it's custom-built to your plant's physics, not configured from a template. It optimizes the entire chain in one model, purchasing, production, inter-plant transfer, distribution, and shipment, the way a configurable APS does. But because the model is built for your plant rather than bent from a discrete template, the changeovers, yields, blends, and co-products live inside the math instead of in a spreadsheet. And because it's one focused custom model rather than a configured enterprise suite, the pilot is in the low five figures, not the high six. End to end like an APS, built for your physics like nothing off the shelf, at a cost that doesn't need a board vote.

ERP plus spreadsheets has no optimization at all. A scheduler optimizes only the floor. A configurable APS optimizes end to end but is built on a discrete template and priced for enterprises. WonForge is a different kind of APS: end-to-end optimization, custom-built to process physics, at mid-market cost.

Frequently Asked Questions

We already have a scheduler. Isn't that enough?

A scheduler sequences the floor, which is real value, but it optimizes one slice in isolation. It takes the demand it's handed and orders the jobs, without seeing purchasing, inter-plant transfers, multi-site capacity, or distribution, and without asking whether that demand mix was the right thing to build. You get a sharp schedule for a plan nobody optimized end to end. WonForge plans the whole chain and sequences the floor in one model, so the schedule reflects the best plan, not just the best order for a fixed one.

How is this different from a configurable APS like the big suites?

WonForge is an APS, so the difference isn't the category, it's how the model is built. Configurable APS suites are configured from a data model built for discrete manufacturing, so process physics like sequence-dependent changeovers, variable yields, and co-products get approximated or pushed into spreadsheets, and the configuration itself is what runs six or seven figures over a project of many months to years. WonForge is custom-built to your physics and is one focused model, so the physics lives in the math instead of in a spreadsheet.

We mostly plan in spreadsheets today. Is that really a problem?

It's the most common setup in mid-market process manufacturing, and it works until it doesn't, when the planner who owns the spreadsheet is out, when SKU count grows, or when the constraints interact in ways no one can solve by hand. The ERP nets requirements but doesn't optimize; the spreadsheet optimizes one piece at a time. The gap between that and a plan that solves the whole thing at once is margin that leaks every month.

Conclusion

Every process plant already plans with something, an ERP and spreadsheets at the base, often a scheduler on top, sometimes a configurable APS. Each handles a piece. The ERP records but doesn't optimize. The scheduler optimizes the floor but not the plan. The configurable APS optimizes end to end but was built on a discrete template and priced for enterprises. The missing option isn't a different category, it's a different kind of APS: end-to-end optimization, custom-built to process physics, at a cost the mid-market can approve. That's the plant WonForge was built for.

Scheduler, ERP, or Configurable APS: Where Each One Stops Short for Process Plants