DELMIA Quintiq Alternatives: What Process Manufacturers Should Evaluate

A grade change takes effect in six weeks. The planning rule that handles it sits in a queue behind two other requests, so a planner builds the workaround in a spreadsheet. Everyone agrees that it’s temporary.

The forecast lives in a different file from the plan it drives. Someone reconciles the two on Thursday afternoons. But that reconciliation isn’t in anyone’s job description.

Only one person understands how the model behaves at the edges. Nobody has said out loud what happens when that person leaves.

None of this reads as a crisis. It’s just the ordinary operating condition of a planning platform that has drifted from the business it models.

If your plants run on DELMIA Quintiq, the platform probably runs well where it was designed to. The question worth asking is what happened to everything around it: Demand planning, forecast collaboration, inventory policy, and the ability to change a planning rule without a development cycle. That work just grew up in spreadsheets alongside the platform. Nobody decided against it.

What DELMIA Quintiq Was Built to Do

Quintiq was founded in the Netherlands in 1997 as an optimization and scheduling engine, and Dassault Systèmes acquired it in 2014. One modeling approach, capable of describing crew rosters, production sequences, and transport networks alike.

The public record of European manufacturing deployments is consistent. Quintiq went into these plants to sequence production, including hot mill and cold mill scheduling; mixing, dispersing, and filling; and melt and cast operations. That is hard work, and Quintiq does it. The gap opens upstream, because a platform designed for breadth arrives without a point of view about your industry.

Where the Planning Model Lives

Quintiq is configured in Quill, Dassault’s proprietary configuration language. Everything exists as model code: your grades, changeover penalties, sequencing rules, and sourcing logic. The downside is where the ownership lies. A change to your planning logic is a change to the model, and model changes are developer work, performed by certified specialists or an implementation partner.

That arrangement makes sense when the model is stable, but process manufacturing models are rarely stable. New grades arrive, campaign structures shift, a shelf-life rule picks up three exceptions, a supplier constraint changes the sourcing logic. Each one is a request in a queue.

This effect compounds. A model built in 2011 has absorbed a decade of change requests, and each was probably reasonable on its own. Together they describe a business that has moved on, encoded by people who have often moved on as well.

Your planners already manage those constraints, usually informally. The open question is whether the platform lets them encode what they already know.

Where Demand Planning Sits

Quintiq offers a demand planning module with statistical and collaborative forecasting. It arrived later than the optimization and scheduling core, as a separate module.

A separate module means a handoff. The forecast gets built in one place, the supply plan in another. Somebody reconciles the two by hand, against a deadline. When the numbers disagree, the discussion turns to which file is current.

That leaves a big question to answer honestly about your own environment: Is your forecast a first-class part of the same planning model that drives supply and production? Or does it live next to it and get handed across?

Arkieva runs demand, inventory, and supply planning in one model. The forecast and the plan it drives are not separated by a handoff.

Forecasting is statistical and collaborative in the same place. Planners work in an interface built for planners, so building and adjusting a forecast is not developer work.

Demand sensing runs as a native module. It takes in short-term signals, including point of sale, orders, and shipments, and refines the near-term forecast against them. A monthly statistical forecast cannot see what happened last week. For a manufacturer with long production lead times, that near-term horizon is where service gets decided.

Promotion planning is built in and integrated with the demand planning engine, including trade promotion planning. A promotion is planned in the same engine that builds the forecast, so the supply plan sees it as part of demand.

Where the Platform Runs

Dassault’s own 2025 and 2026 release documentation specifies that Quintiq is an installed product. It requires sixty-four-bit Windows or Windows Server, administrator rights, and an embedded Java runtime. Dassault also offers managed hosting, where a team operates that installation for you, and its cloud-native scheduling application is a separate, newer product on the 3DEXPERIENCE platform.

It’s worth knowing which of those you are on. The answer determines who carries the operating burden: the infrastructure, the upgrade projects, and the patching windows.

What the Install Base Suggests

Published Quintiq customer stories across European manufacturing concentrate in metals, logistics, aviation, and rail. Steel, aluminium, copper, and zinc appear repeatedly. Chemicals, food and beverage, and CPG appear far less often.

That’s a reasonable outcome for a general optimization engine. It’s also a signal to three areas of concentration: the reference implementations, the accumulated modeling patterns, and the partner expertise.

Your constraints might include shelf life, allergen and changeover rules, co-products, blending, or campaign runs against variable yield. If so, ask how much of that arrived in your model as configuration work your team paid to build.

What Planner Ownership Changes

Arkieva ships process manufacturing constraints as capability. Yield variability, campaign sequencing, shelf life, blending, and multi-stage production are modeled as part of the platform.

It’s built for planners to configure and extend directly. That keeps planners in the driver’s seat, and it keeps planning knowledge in the platform where the team can reach it.

The platform quantifies risk as the plan is built, so the gaps are known before execution. A planner can trace a plan back to the assumptions behind it and show management what the decision was based on.

Six Questions Worth Asking

Each of these points is a decision already being made somewhere in your organization, whether or not anyone has named it as one.

• Who can change a planning rule? How long does that take from request to live?
• Does demand planning sit in the same model as supply and production, or alongside it?
• Are yield variability, grade transitions, and campaign sequencing modeled natively, or handled by convention and planner correction?
• Where does the platform run, and who carries the operating burden?
• When the person who understands the model leaves, what stays behind?
• Across the last five years, how much of your planning spend went to new capability and how much to rework?

A Different Way to Think About the Decision

A scheduling engine that models your plant well is worth having. The question is whether the planning work around it keeps expanding through change requests, or whether the platform lets your team encode what it already knows.

That changes what a change costs. When a planner encodes a new rule directly, the queue stops limiting how fast planning keeps up with the business. Knowledge that lives in the platform also outlasts the person who put it there.

If your planning involves that kind of complexity, it is worth seeing what a platform built specifically for it looks like.

See how a purpose-built alternative compares.