Changeover Optimizer
Assembly Changeover Optimizer
From a work-order spreadsheet to an optimised, changeover-minimised schedule in minutes.
AI-native discrete event simulation
Simulate the decision before it costs you the quarter.
Open Netrikkan turns changeovers, schedules and quotes into evidence. Model your plant, run thousands of scenarios against your real constraints, and get a ranked recommendation with the reasoning visible.
- Nine decision engines
- Sits beside ERP, MES & WMS
- Integrates via APIs, OPC UA & MQTT
Built in BengaluruA UTS CompanyNASSCOM DeepTech memberPatent-pending IPBengaluru, India
The decision layer
Contract manufacturers are paid for decisions made correctly
Which job runs next, which machine takes it, how much stock stands behind it — thousands of times a week. Most are still made on spreadsheets and experience. Simulation lets a plant test a decision before it commits to it.
Scenarios, not forecasts
What if a customer adds 20% volume, a machine is down for two shifts, or a mould is delayed? Each becomes a scenario, compared side by side.
Recommendations with reasons
An AI decision layer ranks the fix and explains why — every scenario is a reproducible run that can be defended to a customer or an auditor.
Additive and private
Reads from ERP, MES and WMS; never replaces them. Language models run locally, so operational data need not leave your environment.
How it works
From event model to recommendation
A four-agent decision chain runs locally against your plant’s own data, in a closed loop that keeps refining the recommendation until improvement flattens.
01
→Interpreter
Reads plant data and constraints from the spreadsheets, PDFs and drawings you already hold.
02
→Diagnostician
Finds where time is being lost — queues, changeovers, idle hours, late material.
03
→Simulator
Runs hundreds or thousands of scenarios and shows the range of outcomes, not one average.
04
Recommender
Ranks the fix, explains why, and hands the plan to the planner.
↻ Closed loop — repeats until improvement flattens
Modules
Nine decision engines. One question each.
Start with the decision that costs the most today. Each module runs on the same simulation core and can be adopted on its own.
01High-mix assembly
Changeover Optimizer
Which sequence of work orders cuts changeover time this week?
Available moduleExplore
02Auto components
Newline
Which equipment and layout does a new component line need?
Available moduleExplore
03Any machine bank
Machine Bank Scheduler
Which job runs on which machine in the bank, on which shift?
Available moduleExplore
04Pharma
Tablet
Where does batch flow stall on the tablet line?
Available moduleExplore05
Distribution / 3PL
Warehouse
Where will inbound and dock operations jam this peak?
Available moduleExplore06
Warehouse outbound
Outbound
Which pick, wave and staffing plan ships on time at lowest cost?
Available moduleExplore07
Semiconductor
FabSim
Where is the fab losing yield, capacity and cycle time?
Available moduleExplore08
Oil & gas upstream
OilSim
What is the probability crude reaches the refinery on time?
Available moduleExplore09
FMCG distribution
Replenish
Which freezers need restocking, and in what order?
Available moduleExploreSee it running
Product walkthroughs
Watch the modules in action — from an order spreadsheet to a verified plan.
Newline
Newline — New Component Feasibility & Line Planner
A blank page to a fully scoped, cost-and-capacity-checked production line.
Tablet
Tablet — Pharma Line Optimizer
Simulate it first: model the tablet line, run the scenario, see the outcome before anything moves on the floor.
Machine Bank Scheduler
Customer Commitment Analyser
Verify the floor before Sales commits: verified delivery dates, machine allocation and labour in one flow.
9decision engines
9integration categories supported
2,000Monte Carlo trials in OilSim
6simulation modules in FabSim
Industries
One method across sectors
Changeover hours, quote accuracy, utilisation, batch lead time, dock cost, yield, stock-out risk, delivery cost — nine cost lines, one method.
Injection moulding & plastics
Machine-bank scheduling and verified delivery commitments.
Automotive components
Quote new lines faster and more accurately.
High-mix assembly
Cut changeover hours without capital spend.
Pharma & life sciences
Batch flow, hold-times and validated scenarios.
Warehousing & 3PL
Dock, slotting and labour planning.
Semiconductor
Yield, capacity and cycle-time levers.
Oil & gas
Probability that crude reaches the refinery on time.
FMCG distribution
Freezer restocking and route priority.
Case studies
Outcomes from our programmes
Anonymised outcomes from digital-twin programmes and our own modules in action.
Aerospace · Digital twinVirtual commissioning of an aircraft-fuselage inspection cell
Prove a robotic scanning process for a full aircraft fuselage in a physics-accurate digital twin before any hardware was commissioned.
Read the outcomes
Aerospace · RoboticsAutonomous scanning robot: motion and collision-safety design
Validate how a scanning robot moves position to position through the fuselage structure and stops precisely without collisions.
Read the outcomes
Pharma · Open NetrikkanValidating a tablet-line configuration before the floor
Find a better line configuration without running a trial batch — the batch that would otherwise cost more than most improvement projects.
Read the outcomesIntegrations
Works with the systems you already run
ERP, MES, WMS, digital twin, historians and shop-floor protocols — and open APIs for anything else.
ERP
- SAP
- Oracle
- Microsoft Dynamics 365
- Infor
- IFS
- Epicor
MES / MOM
- Siemens Opcenter
- DELMIA Apriso
- Plex
- Rockwell FactoryTalk
- AVEVA MES
WMS
- SAP EWM
- Manhattan Associates
- Blue Yonder
- Oracle WMS
- Körber
- Infor WMS
Digital twin & 3D
- NVIDIA Omniverse
- OpenUSD
- Siemens Tecnomatix
- Azure Digital Twins
- AWS IoT TwinMaker
Historians
- AVEVA PI
- AVEVA Historian
- GE Proficy
- Ignition
- InfluxDB
Protocols & standards
- OPC UA
- MQTT
- Sparkplug
- Modbus
- Apache Kafka
- REST / OpenAPI
- Webhooks
See it against your own numbers
A 30-minute working session: you nominate the planning decision that hurts most, and we run the closest tool against a representative case from your operation.