Plan, shop floor and quality in one flow
In most companies manufacturing lives in two places: an ERP that knows the cost and a spreadsheet that knows the plan. As a result the product cost arrives at month end, approximately, and the question “why was this order late” is answered from a planner's memory rather than from the system. NG One joins engineering data, planning, shop floor execution, quality control and maintenance into one flow over the same data. Every material issue, labour hour and downtime event carries plant, line, order and project dimensions from the moment it happens. Planning follows the UX proven in NG Operations — the grid, the planner cockpit and the what-if scenario.
- BOMs and routings
- MRP and planning
- Work orders
- Production Grid
- QMS and lab
- Projects and timesheets
Manufacturing without the blind spot between plan and order
This space covers the path from a bill of material to a finished product, and from a project to a billable hour — including quality and maintenance, which domestic systems typically lack.
The problem rarely starts on the shop floor. It starts when the bill of material lives in a spreadsheet, the consumption norm lives in an engineer's head, the plan lives in another spreadsheet, and production confirmation arrives in accounting on paper five days later. The system then knows what was produced, but not when, on which machine, with what downtime and at what cost. When a customer asks why an order was late, or why a batch cost more than quoted, the answer is reconstructed by hand — and it usually disagrees with the accounting entries.
NG One treats that chain as one document flow rather than five disconnected records. The work order, the material requisition, the production confirmation, the inspection order and the maintenance order all run on the same document framework: a state machine with named guards, five dates on the document, approvals and maker–checker from the workflow kernel, and an immutable change log. A confirmation from a shop floor terminal is a transaction at the moment it happens, not a form someone retypes later. Consumption, labour and scrap carry dimensions immediately, because a dimension is part of the posting rather than a field added when a report is needed.
The structural advantage comes from two decisions that cannot be retrofitted. First: dimensions and effective-dated rules sit in the core rather than in a reporting layer, so order, line and project profitability is read from the ledger instead of being assembled in a spreadsheet. Second: planning is not reinvented — the grid, the planner cockpit and what-if are taken from NG Operations, a product that already does this work in complex manufacturing, and which NG One orchestrates when the plant is more complex than an ERP should carry. On top of that, quality, lab and maintenance are part of the same core and the same audit trail, not bolt-on modules with their own database.
The flow through this space
Plan-to-Fulfill as it actually runs in a company — from engineering data to order settlement and posting with dimensions.
- Step 1
Engineering data
Before anything is planned, the product must be described so that planning, the shop floor and accounting all compute from the same definition.
- Bill of material with version and effective date
- Routing, operations and work centres
- Consumption norm, allowance and expected scrap
- Standard cost estimate
- Step 2
Plan and MRP
The sales and production plan feeds MRP, which derives net from gross requirements and proposes what to make and what to buy.
- Sales and production plan by period
- MRP: gross requirements → net requirements → proposals
- Proposed work orders and purchase orders
- Material coverage check before release
- Step 3
Scheduling
A proposal is not a plan until it passes through capacity. The planner sees the load, moves an order and sees the consequence before publishing.
- Production Grid with load per work centre
- Planner cockpit: delays, bottlenecks, exceptions
- What-if scenario before the plan is published
- Published plan → work orders released
- Step 4
Shop floor execution
Material is issued against the order and production is confirmed where it happens — on a terminal, at the moment it happens.
- Requisition and material issue per order
- Confirmation of output, consumption and labour
- Downtime recorded with a reason code
- Traceability: material lot → product lot
- Step 5
Quality
Inspection is a step in the flow, not paperwork afterwards: an order does not move on until the control point has a decision.
- Inspection plan and sampling per operation
- Lab inspection order and results
- Non-conformance → quarantine → disposition
- Certificate and declaration with the product lot
- Step 6
Settlement and close
The order closes with an actual cost calculation, and the variance is read by cause — material, labour, scrap or subcontracting.
- Actual costing of the work order
- Variance against the standard cost by cause
- Finished goods receipt into stock
- Posting with plant, order and project dimensions
Capabilities
Thirty named capabilities in five groups — from the bill of material and the routing through to actual costing, quality, maintenance and project profitability. All of it runs over the same data, in the same audit trail, with the same dimensions.
Engineering data
One versioned, effective-dated product definition that planning, the shop floor and accounting all compute from.
6 capabilities
Items, UoM conversions and barcodes
One item record with unit-of-measure conversions and barcodes. The foundation the bill of material, the shop floor and accounting all compute from, instead of each keeping its own code for the same thing.
Bills of material
Multi-level BOMs with version, effective date and alternative materials. Changing a BOM does not rewrite the history of orders already produced.
Routings and work centres
Operations with setup, run and queue times, tied to work centres, machines and capacity calendars. The routing feeds both costing and scheduling.
Norms, allowance and expected scrap
Standard consumption per operation, permitted allowance and expected scrap. Without a norm a variance has no reference — and no answer to why a batch cost more.
Standard cost estimate
Product cost derived from the BOM, the routing and current material prices before the order starts. Later it serves as the reference for variance.
PLM and engineering change
Product lifecycle and controlled engineering change orders: who requested the change, who approved it and from which date it applies to the BOM.
Planning, MRP and scheduling
From the sales plan to a published schedule. The planning surface follows the UX proven in NG Operations (ngoperations.app).
7 capabilities
Sales and production plan
A plan by period, item and group, with plan versus actual. An input to MRP, not a separate document nobody opens.
MRP
Gross requirements from the plan and open orders, net requirements after stock, reservations and inbound. Output is proposed work and purchase orders, each showing why it exists.
Production Grid
A dense planning table showing load per work centre and period, where moving an order recalculates the consequence immediately.
Planner cockpit
One screen with what needs a decision: delays, bottlenecks, uncovered material, exceeded capacity. Exceptions rather than generic charts.
What-if scenarios
A copy of the plan where a move, a priority change or an extra shift can be tried without touching the published plan until the scenario is accepted.
APS scheduling
Finite scheduling constrained by capacity, tooling, people and calendars. The engine proposes the schedule and explains why an order landed on that slot.
Demand forecasting
A proposed plan derived from sales history and seasonality, with drill-down to the periods and items it was derived from. The planner accepts, edits or rejects it.
Shop floor execution
The work order from release to close, confirmed where the work actually happens.
6 capabilities
Work orders
Created from an MRP proposal or manually, released to the floor, with a state machine of permitted transitions and approvals from the workflow kernel.
Requisition and material issue
Requisition per order or operation, issue with lot and serial, return of unused material. Stock changes at the same moment.
Confirmation of output, consumption and labour
Good quantity, scrap, consumed material and labour time confirmed per operation, with downtime recorded against a reason code.
Shop floor terminals
A screen built for gloves and dirty hands: large targets, few fields, no mouse required. A floor operator is a licence granule, not a full named user.
Subcontracting
An operation performed by a third party: send and receive, quantity held at the subcontractor, and a service cost that lands in the order calculation.
Actual costing of the work order
Real product cost from consumption, labour, scrap and subcontracting, with the variance against the standard broken down by cause.
Quality, lab and maintenance
QMS, lab and CMMS inside the same core and the same audit trail — not bolt-on modules with their own database.
6 capabilities
Inspection plans and sampling
Control points at goods receipt, in process and at output, with a sampling plan and measured characteristics. The order does not move on until the point has a decision.
Non-conformance and corrective action
A non-conformance record, quarantine, a disposition (rework, scrap, concession) and a corrective action with an owner and a due date.
Laboratory
Inspection orders, methods, limits and results, producing the certificate and declaration issued with the finished product lot.
Lot-to-lot traceability
Batch genealogy in both directions: from a finished lot down to every material lot, and from a suspect material up to every customer who received it.
Equipment maintenance (CMMS)
Equipment register, preventive plans by time or output, maintenance work orders, spare parts, and downtime linked to the production order it affected.
Anomaly detection on the floor
Consumption drifting from the norm, rising scrap or a stretched operation time raise an exception with evidence — which order, which operation, how large the deviation.
Projects, resources and profitability
Project-to-Cash for companies that deliver a project rather than a series — with profitability read from the ledger.
5 capabilities
Projects and structure
A project with phases and a work breakdown structure, deadlines, owners and links to every document raised against it.
Tasks and resources
Tasks on the workflow kernel, scheduling of people and equipment, resource load and visible booking conflicts.
Timesheets
Hours by project, phase and task, approved by a manager and split into billable and non-billable.
Project budget and controlling
Budget by phase and cost type, plan versus actual during the project, and a warning before an overrun rather than after it.
Project profitability
Revenue, direct cost, hours and margin per project — from dimensions on the posting, with no spreadsheet assembly afterwards.
What AI does in this space
One rule holds throughout: the output is a draft or an exception with evidence, and a person confirms the decision. AI works inside the user's own permissions and tenant isolation, and every suggestion enters the AI audit trail.
Demand forecasting as an input to the plan
The model learns from sales history, seasonality and the calendar, and proposes a plan by item and period. Every number carries a drill-down to the data it was derived from and the range it moves in. The planner accepts, edits or rejects — the plan never publishes itself.
APS scheduling that explains itself
The scheduling engine takes capacity, calendars, tooling, people and priorities and returns a schedule. With it comes the answer to why an order landed on that slot and which constraint was the bottleneck. Not a black box: the constraints are configured and visible.
Anomaly detection on the shop floor
The system watches consumption drifting from the norm, scrap rising on an operation, run times stretching and repeated downtime on the same machine. A finding never blocks work — it raises an exception on My Work with the order, the operation and the size of the deviation. Thresholds are set per plant.
Copilot on the work order
Ask “why did this order cost more than the estimate” and the answer arrives with source documents: which requisition, which confirmation, which downtime and which subcontracting line make up the difference. Every conclusion drills down to the data. If the data is not there, the Copilot says so instead of guessing.
Drafts instead of silent automation
When MRP reports uncovered material, or inspection returns a non-conformance, NG One prepares a draft: a purchase order, a requisition, a task for the owner or a message to the subcontractor. The draft waits for confirmation and passes the same approval limits as a manually entered document. Agent autonomy is configured per process.
Why this is different
Measured against what companies run today — Pantheon, Business Central, Odoo or a legacy system propped up by spreadsheets.
Dimensions from day one, not a field added later
Plant, line, work order and project are dimensions inside the posting itself, from the first line. Order and project profitability is therefore read from the general ledger instead of being assembled from an export. Systems that add a dimension later can show it on a document, but not in history that is already posted.
Planning built on a UX that already runs a plant
The grid, the planner cockpit and what-if were not invented for this ERP — they come from NG Operations (ngoperations.app), a product that plans complex manufacturing. NG One carries the manufacturing an ERP should carry; NG Operations takes over when the plant is more complex than that. The boundary is explicit rather than a third tab on the work order.
Backdated confirmation without a wrong product cost
Confirmation posted after the fact is the rule, not the exception — paper from the floor is always late. NG One recalculates stock and costing backwards instead of demanding a chain of reversals or leaving a product cost that will not reconcile with the stock ledger. That decision lives in the inventory core and is available to manufacturing from the first order.
Quality and maintenance in the core, not alongside it
QMS, lab and CMMS are the largest gap in domestic systems and are usually solved by a separate program with its own database. Here they are the same document framework, the same approval workflow, the same audit trail and the same dimensions. A non-conformance links the lot, the order, the machine and the customer without a single export.
The flows this space runs through
A business space is not an island. These processes touch it end to end, and where a flow leaves this space the record stays the same — the next step receives it structured rather than retyped.
Operations
Plan-to-Fulfill
The path from plan to delivered product. The plan explodes through the bill of materials into material requirements, and actuals from the floor and the warehouse return to the order as it runs. The warehouse is a real WMS: bin locations, directed picking, waves.
- Plan
- Material sourcing
- Production
- Quality
- Stock
- Delivery
Projects
Project-to-Cash
The path from a signed engagement to a settled progress claim. Hours, materials and subcontracts charge the same project, so margin is visible while the work runs rather than after it.
- Engagement
- Project and budget
- Tasks and resources
- Timesheet
- Progress claim and invoice
- Profitability
Questions about this space
Scope, boundaries, and the rules this space posts by.
Does NG One replace NG Operations?
No. NG Operations is a separate product for planning complex manufacturing, and NG One orchestrates it rather than absorbing it. NG One carries the manufacturing an ERP should carry: BOMs, routings, MRP, work orders, confirmation, costing and quality. When the plant is more complex — many constraints, scheduling around tooling and people, frequent re-planning — planning stays in NG Operations while NG One holds master data, material, cost and posting. The planning UX in NG One (grid, cockpit, what-if) is taken from that product precisely because it is proven on a real shop floor. More at ngoperations.app.
In what order does manufacturing get rolled out at our site?
A plant does not move over a single weekend, so the rollout runs in project phases. Master data first: items and UoM conversions, BOMs with version and effective date, routings, work centres and norms — without them neither planning nor costing has anything to compute from. Then the work order with requisition, confirmation and actual costing, then the terminals on the floor, then quality and maintenance. Planning — the grid, the cockpit and what-if — goes live once master data can carry a plan, because scheduling on wrong operation times produces a tidy schedule that does not hold. The order is driven by where the company loses the most today, not by the order of the screens in the menu.
Does a shop floor operator need a full user licence?
No. A shop floor terminal is a licence granule, separate from a full named user licence. The licensing and entitlement framework sits in the core for exactly this reason: packages, licences by user type and feature entitlements are part of the platform rather than an afterthought on a price list, because a licensing model cannot be retrofitted painlessly. In practice a shift of thirty people confirming production does not cost the same as thirty ERP users.
How is the real product cost produced?
In two steps. The standard cost estimate is derived from the BOM, the routing and current material prices before the order starts, and serves as the reference. Actual costing then sums real consumption, confirmed labour, scrap and subcontracting, and breaks the variance down by cause rather than showing one lump figure. Because confirmation from the floor often arrives late, the system recalculates backwards — the product cost is corrected instead of staying wrong until the next stock count.
We run process manufacturing with batches and certificates. Is that covered?
Yes, and that is why the lab and traceability sit in the same space as the work order. Material arrives with a lot, an expiry date and FEFO logic from the inventory core, testing runs as an inspection order with methods and limits, and the certificate and declaration are issued with the finished product lot. Genealogy works in both directions: from a finished batch down to every input lot, and from a suspect material up to every customer who received it. When an inspector or a customer asks for the batch file, it comes out of the system instead of being reconstructed from three programs and a binder.
Can we use projects without manufacturing?
Yes. Projects, tasks and resources, timesheets, budget and profitability are a separate flow — Project-to-Cash — and make sense for engineering, installation, field service and professional services without a single work order. Timesheet hours and costs carry the project dimension into the posting, so margin per project is visible during delivery rather than after invoicing. Project invoicing and advances canonically belong to the Sales and customers space, but are fully reachable from the project.
We use subcontractors for some operations. How is that handled?
Subcontracting is an operation in the routing like any other, except the performer is a third party. Dispatch and receipt are tracked, along with the quantity held at the subcontractor and the due date, and the service cost lands in the work order calculation instead of ending up as overhead. The subcontractor's incoming invoice goes through 3-way matching like any other purchase, so the gap between the agreed and the invoiced service is visible before payment.
Neighbouring spaces
Show us your plant, not your requirements list
The most useful conversation starts from one real bill of material, one order that ran late and one batch that cost more than planned. That is enough to show how NG One carries your product through planning, the floor and costing, where the variance breaks down by cause, and where the boundary towards NG Operations runs.