Composable, app-based manufacturing platforms have kept building momentum through 2025 and into 2026. Tulip Interfaces, the best-known vendor in this “no-code MES” category, has continued adding funding, systems-integrator relationships, and hardware/software partnerships aimed at making frontline-ops apps easier to deploy alongside existing plant systems. Competing platforms — from newer app-builder entrants to established suites bolting on low-code app layers — are chasing the same wedge: let a process engineer or continuous-improvement lead build a work-instruction, andon, or line-clearance app in days instead of commissioning a change request against a traditional MES.
That momentum is now showing up where it counts — in actual RFPs. Plant IT and manufacturing engineering teams that once treated MES as a two- or three-vendor conversation (think Siemens Opcenter, Rockwell/Plex, GE Digital’s Proficy, or a homegrown SAP-adjacent system) are now fielding serious proposals from composable platforms as either a replacement or a companion. That’s a meaningfully different conversation than it was even a couple of years ago, and it’s one a lot of buying committees aren’t yet equipped to have cleanly.
What “composable MES” actually means
Strip away the marketing language and composable MES platforms share a few real characteristics: a drag-and-drop or low-code app builder, a library of connectivity widgets for PLCs, scales, scanners, and vision systems, a device-agnostic runtime (tablets, HMI panels, wearables), and a cloud or hybrid data layer that’s more flexible than the rigid data models baked into traditional MES suites. The pitch is that a manufacturing engineer — not a systems integrator — can stand up a digital work instruction, an andon escalation flow, or a manual SPC data-collection app in a sprint, then iterate on it the same week based on operator feedback.
Traditional MES suites, by contrast, are built around ISA-95 data models, tight ISA-88 batch/recipe structures where relevant, and deep integration with ERP for order execution, genealogy, and compliance record-keeping. That architecture is slower to configure but it’s also the reason regulated manufacturers lean on it: the data model enforces consistency, traceability, and audit trails as a structural property of the system, not as something an app builder happens to produce correctly.
Where composable platforms genuinely replace MES functionality
There’s a real, defensible set of use cases where a platform like Tulip can substitute for — not just supplement — a legacy MES module:
- Digital work instructions and visual line-side guidance, especially where instructions change frequently or vary by SKU and the traditional MES’s document-control workflow is too slow to keep up.
- Andon and escalation management, where the value is fast configuration of triggers, notification chains, and dashboards rather than deep integration with scheduling logic.
- Kitting and manual assembly verification, particularly pick-to-light or vision-assisted kitting steps that legacy MES modules handle clumsily or not at all.
- Manual SPC and quality data collection at stations that aren’t natively instrumented, where the alternative is a paper traveler or an Excel macro nobody trusts.
- Rapid pilot cells and new-product-introduction lines, where the process is still being defined and a rigid MES configuration cycle would lag the engineering changes.
In all of these, the common thread is that the process is operator-facing, changes often, and doesn’t require the app itself to be the authoritative genealogy or batch record.
Where you still need a system of record
Where composable platforms are not yet a substitute — and, in our assessment, won’t be for most regulated or high-complexity discrete/process environments any time soon — is anywhere the software has to be the authoritative source for scheduling, detailed genealogy, or regulatory batch records. Finite scheduling against constrained resources, full material genealogy tied to ERP-driven lot and serial control, electronic batch records under GMP, and complex recipe management with parameter-level exception handling are all places where the data model rigor of an Opcenter, Plex, or Proficy earns its keep. Composable platforms can feed data into those systems of record beautifully. Asking them to become the system of record is a different and much riskier bet, and it’s usually not what these vendors are actually pitching, whatever an enthusiastic salesperson in an RFP meeting might imply.
The scoping checklist: how to pilot this without creating shadow IT
The failure mode nobody puts in the RFP is the quiet one: a composable-MES pilot succeeds on one line, gets copied by three more supervisors who liked what they saw, and eighteen months later you’ve got a dozen semi-related apps holding production data that IT didn’t provision, security didn’t review, and quality can’t audit. That’s shadow IT with a nicer UI. Here’s how to scope a pilot so that doesn’t happen:
- Name the system of record before you build anything. Decide explicitly which system owns genealogy, order execution, and compliance records, and treat the composable app as a client to that system, not a peer.
- Define the integration contract up front. Specify what data flows to your historian, ERP, or MES of record, over what protocol (OPC UA, MQTT Sparkplug B, REST), on what cadence, and who owns the mapping when tag names or schemas change.
- Put IT and security in the pilot from day one, not at go-live. Composable platforms often run cloud-connected; that means IEC 62443 zone/conduit thinking, data residency questions, and identity/access management need to be settled before line-side deployment, not retrofitted after.
- Set an app governance model before app number two gets built. Decide who can publish an app to production, how versions are controlled, and how an app gets retired — the same discipline you’d apply to any other production software, even though the tool makes it feel like a spreadsheet.
- Scope the pilot to a bounded, non-regulated process — a kitting station, an andon flow, a pilot cell — with a defined success metric and an explicit decision point for whether it scales, integrates deeper, or gets retired.
- Ask the vendor directly what they don’t claim to do. A vendor that’s candid about needing to sit alongside a traditional MES for scheduling and genealogy is being straight with you; one that implies full replacement for a regulated batch process deserves more scrutiny, not less.
What to watch through the rest of 2026
The signal worth tracking isn’t another funding round or another integrator partnership announcement — those are lagging indicators of investor and channel interest, not proof of fit for your plant. The real signal is whether composable platforms start showing credible, audited integrations into genealogy and batch-record systems rather than just data export. That’s the gap that currently separates “great app layer for the frontline” from “legitimate MES alternative” in regulated environments. Until you see that maturity demonstrated at a customer reference you can actually call, the sound buying posture is to treat composable MES as a fast, flexible complement to your system of record — and to scope every pilot with an exit plan, not just an entry point.
This article was written with the assistance of artificial intelligence. While we aim for accuracy, the information may be incomplete, out of date, or incorrect, and should be independently verified before you rely on it for any decision. It is provided for general information only and does not constitute professional advice.
