The Challenge: The Paradox of Too Much and Too Little
In defence manufacturing, a Bill of Materials is not a flat list. It is a tree — sometimes eight or ten levels deep — where each node represents a sub-assembly that must be fully kitted before the assembly above it can begin. A top-level assembly might require hundreds of child parts, some of which are themselves assemblies requiring dozens of components. Miss one leaf node on this tree and the entire branch is blocked.
For a leading engineering conglomerate operating in this space, the challenge was precisely this: the traditional MRP-based planning system they operated could not identify which missing part was actually blocking production. MRP is designed to ensure that every part is ordered when needed. It does this well, in theory. In practice, forecast inaccuracies, supplier delays, and engineering revision cycles mean that some parts arrive on time, others arrive late, and MRP has no mechanism for prioritising which shortage matters most.
The result was the classic inventory paradox: simultaneously too much and too little. Too much of the parts that had arrived on schedule, piling up as work-in-progress waiting for their missing companions. Too little of the specific components that were actually blocking assembly. Stores held significant inventory value, but assembly lines stopped anyway. Production planners spent their days firefighting — chasing individual part shortages through phone calls and emails — rather than planning.
Engineering revision changes added another layer of complexity. In defence manufacturing, design changes are frequent and non-trivial. A component revision can cascade through multiple levels of a hierarchical BOM. Under the previous system, tracking which work orders were affected by a given engineering change required manual BOM interrogation that could take days. Occasionally, work orders were released against obsolete BOMs, creating scrap and rework.
The Theory of Constraints Approach: Constraint-First, Not Volume-First
The intellectual foundation of the FlowTrack MICOS implementation is Goldratt's Theory of Constraints applied to material planning. Where MRP asks "what should we order for all assemblies?", TOC-based planning asks "what is blocking the highest-value assembly right now, and how do we remove that specific constraint?"
The full-kit system operationalises this at the work order level. Before a work order is released to the shop floor, MICOS performs a full-kit check: it traverses the entire BOM tree for that assembly — all levels, all components — and verifies that every required part is physically available in the correct quantity. If any component is missing, the work order is held. It does not start.
This sounds simple. The implications are profound. Under the previous approach, a work order would be released when MRP calculated that it should start, regardless of whether all components were actually present. Workers would begin the assembly, reach the missing component, and stop. The partially assembled unit would sit as work-in-progress — occupying space, tying up capital, and appearing on the production schedule as "in progress" when it was actually blocked. The real shortage was invisible until someone physically walked the shop floor.
Under the full-kit model, the shortage is visible before the work order starts. The planner can see exactly which component is missing, what the expected arrival date is, and which other work orders are also blocked by the same shortage. Resources can be redirected to assemblies that are fully kitted. The constraint — the single missing part blocking the most valuable assembly — is the explicit focus of expediting effort, not discovered by accident on the shop floor.
The shift from MRP to full-kit planning is a shift from "we ordered everything on time" to "we know exactly what is blocking us and we are working on that specifically." These are very different operating states.
PLM and ERP Integration: Live BOM, Live Availability
The full-kit check is only as good as the data it runs against. Two data feeds are critical: the engineering BOM and the material availability position. MICOS integrates with both source systems directly.
Hierarchical BOM data flows from the PLM system via a live interface. When an engineering revision is released and approved, the BOM update propagates automatically to MICOS. The system re-evaluates full-kit status for all open work orders affected by the change. Work orders that were previously green (fully kitted) may shift to blocked if the revision introduces a new component requirement. Work orders that were blocked may become green if the revision removes a component that had been the missing item. The planner sees the current state, not yesterday's state.
Material availability — on-hand stock, allocated quantities, pending receipts by expected date — is drawn from the ERP system in real time. The MICOS availability plan shows, for each blocked work order, which components are missing and when they are expected based on open purchase orders. This allows planners to make informed decisions: is the wait for a specific component measured in days or weeks? Is there an alternative source? Is there a higher-priority assembly that can be completed while waiting?
The system also handles the hierarchical complexity of multi-level BOMs explicitly. A full-kit check at level one triggers sub-kit checks at every child assembly level. A sub-assembly that itself requires components from stores is checked against current availability, not theoretical requirements. This prevents a common failure mode in MRP implementations: a work order released as "fully kitted" at the top level, where the kit check only verified that the sub-assembly was available — without checking whether the sub-assembly itself was actually ready to be built.
Outcomes: Fewer Stoppages, Lower WIP, Data-Driven Release
The measurable headline outcome — a 35% reduction in assembly line stoppages attributable to missing parts — understates the operational change. The more significant shift is structural: the production planning process is now data-driven rather than intuition-driven.
Work order release decisions are made against a system-calculated full-kit status, not a planner's best guess about what is in stores. If the kit is not full, the order does not start. This single discipline produces several downstream benefits:
- Reduced WIP: Partially assembled units no longer accumulate on the shop floor waiting for missing components. Work that starts can finish.
- Focused expediting: The shortage report identifies the specific components blocking the highest-value assemblies. Expediting effort is concentrated on the actual constraint, not spread across all open shortages.
- Inventory rationalisation: When planning focuses on constraints, excess inventory of non-constraint parts is reduced over time. Purchasing decisions are driven by what actually blocks production, not by safety-stock formulas applied uniformly.
- Engineering change management: Revision impacts are automatically propagated to open work orders. The planner sees which orders are affected immediately, without manual BOM interrogation.
The deployment includes a handheld barcode terminal client, which means store-keepers and shop floor supervisors can perform real-time kit checks and material transactions without returning to a workstation. The availability plan updates as transactions are posted, keeping the full-kit status current throughout the shift.
For defence manufacturing organisations where a line stoppage is not merely a production loss but a potential contract milestone breach, the argument for constraint-aware planning over traditional MRP is not theoretical. The 35% reduction in stoppages translates directly to schedule adherence, and schedule adherence in defence contracts translates directly to commercial and reputational standing.