Capabilities
What you can model
180 capabilities, each shown by a test that builds and runs a factory, and 18 guides on how to model a production principle.
Every line below is shown by a test that builds and runs a model. Where a guide explains how to model it, the line links to it. Capabilities that work only in part, or not yet, are left out until they do.
Modelling guides
Built in
Batch machines: furnaces, ovens, autoclaves
A machine processes several units together: they are loaded, the cycle runs once for all of them, and they come out together (heat treatment, autoclave curing, a wash basket, a kiln).
Built in
Breakdowns and repairs
Machines fail at random and are repaired.
Built in
Demand peaks, campaigns, seasons and a mix that changes by shift
Demand is not even over the day, the week or the season: a busy shift, a campaign, a peak month, or a mix that is mostly one product in the morning and another in the evening.
Pattern
Drying, curing and other waiting
A unit has to wait a fixed time between two steps and nothing is worked on it: paint or adhesive drying, curing, cooling, settling, a rest before testing.
Built in
Inspection, sampling and rework loops
A test or inspection sends some units back to an earlier step, or only a share of units is inspected.
Approximation
Loading the next unit while the machine works
A machine has two fixtures, a turntable or a shuttle table: the next unit is loaded while the current one is processed, so loading costs no machine time.
Built in
Machine tending: an operator leaving a machine while it runs
An operator loads a machine, starts it and goes to another machine while it runs, coming back to unload: machine tending, one person on several machines, semi-automatic stations.
Pattern
Make-to-order with several order types
Work starts only for an order, and order types take different routes.
Built in
Make-to-stock with a kanban loop
Products are sold from a finished-goods store, and the factory makes them again as they are sold, so customers do not wait for the lead time.
Pattern
Material used by length, area or weight
A product uses material by measure rather than by piece: metres of cable, square metres of sheet, kilograms of powder.
Built in
Parallel branches that meet: sub-assemblies
Parts of a product are made at the same time on their own machines and meet later: a motor and a frame assembled into a fan, a board and a housing into a module.
Pattern
Products too big for a machine
Some units are too long, wide or heavy for a machine and must go to a bigger one.
Built in
Rush orders
An urgent order arrives while the factory is busy and must overtake the queue.
Built in
Scrap bins, by-products and recycling into raw material
Rejected units go somewhere on the floor, not into thin air: a bin beside the station that rejects them, emptied every so often.
Built in
Scrap, yield and OEE
An inspection rejects a share of units for good (a leak test, a final check), and the factory must still deliver the good units its customers ask for.
Built in
Shifts, breaks and working days
The factory works in shifts, stops for breaks, and is closed at night or at weekends, while some equipment (an oven, a drying rack, a machine running unattended) keeps going.
Built in
Task times from measurements
You have timed a step on the floor and want the model to vary as the real times do, long tail included, without choosing a distribution.
Pattern
Work that repeats a varying number of times
The same work is done several times on one unit, and how many times depends on the unit: holes per plate, welds per frame, screws per cover.
Capabilities, by area
Products and material structure
- One product, one itemSized by Build layoutChecked against a known answer
- Several product familiesSized by Build layout
- Product variants and configurationSized by Build layout
- Multilevel bill of materialsSized by Build layout
- Material quantities per operationSized by Build layout
- Optional componentsSized by Build layout
- Offcuts, by-products and scrap with a flow of their ownHow to model it
- Units, weight and volumeSized by Build layoutHow to model it
- Product dimensionsSized by Build layout
- Serial numbers and unique identity
- Recycling scrap back into raw materialHow to model it
Processes and operation dependencies
- Sequential operationsSized by Build layoutChecked against a known answer
- Parallel operationsSized by Build layoutHow to model it
- Precedence from several branchesSized by Build layout
- Operations without a machineSized by Build layoutHow to model it
- Alternative process stepsSized by Build layoutImprove
- Optional process stepsSized by Build layout
- Repeated operationsSized by Build layoutHow to model it
- Operation time that depends on volumeSized by Build layout
- Assembly joinSized by Build layout
- Consumption and transformationSized by Build layoutChecked against a known answer
Flow topology and networks
- Straight serial lineSized by Build layoutImproveChecked against a known answer
- Joining flowsSized by Build layoutChecked against a known answer
- Optional routingSized by Build layout
- Many-to-many connectionsSized by Build layoutChecked against a known answer
- Re-entrant manufacturing flowSized by Build layoutChecked against a known answer
- Loops with a limited number of rounds
- Rework loopSized by Build layoutHow to model it
- Overlapping independent linesSized by Build layout
- Flow specific to a product familySized by Build layout
- Sub-assemblies that join backSized by Build layout
- Open job shopSized by Build layoutChecked against a known answer
- Scrap and good units leave by different waysHow to model it
Machines and stations
- Single-position machineSized by Build layoutChecked against a known answer
- Parallel identical machinesSized by Build layoutImproveChecked against a known answer
- Parallel non-identical machines
- Multi-operation machineSized by Build layoutChecked against a known answer
- Machines with several working positionsSized by Build layout
- Unattended automatic machineHow to model it
- Semi-automatic stationsHow to model it
- Manual stationsSized by Build layout
- Capability that depends on the tooling set
- Stations with limited infeed and outfeedChecked against a known answer
- Stations with their own bufferSized by Build layout
People and staffing
- One operator per stationSized by Build layout
- Shared operatorsSized by Build layoutImprove
- Several operators for one job
- Manual and automatic shares of a cycleHow to model it
- Operator walking time
- Breaks and shift handoversSized by Build layoutHow to model it
- Staffing as a decision variableSized by Build layoutImprove
- Two-person jobs for ergonomics or safety
Buffers, work in process and storage
- Unlimited queuesChecked against a known answer
- Finite bufferSized by Build layoutChecked against a known answer
- Zero-buffer transferSized by Build layout
- Buffers limited by piecesSized by Build layoutImprove
- Buffers shared by several routes
- FIFO buffersChecked against a known answer
- Priority buffers
- Separate in- and out-buffers
- Work-in-process capChecked against a known answer
- Blocked and starved as separate statesChecked against a known answer
Orders, demand and variation
- One large one-off orderSized by Build layoutHow to model it
- Continuous demandSized by Build layout
- Stochastic order arrivalsChecked against a known answer
- Product mix in proportionsSized by Build layout
- Product mix that changes over timeHow to model it
- Orders with quantity and due dateChecked against a known answerHow to model it
- Orders with priorityHow to model it
- Rush ordersHow to model it
- Seasonal and campaign demand peaksSized by Build layoutHow to model it
- Make-to-stock production with a kanban loopHow to model it
Batches, setups and changeovers
- Fixed batch sizeHow to model it
- Minimum batch sizeHow to model it
- Maximum batch sizeHow to model it
- Batch timeoutHow to model it
- Parallel batch processingHow to model it
Time models and randomness
- Constant process timesSized by Build layoutChecked against a known answer
- Product-dependent process timesSized by Build layout
- Time that depends on length or quantitySized by Build layout
- Triangular time distributionHow to model it
- Normal and lognormal timesHow to model it
- Empirical time distribution from measured timesHow to model it
- Labour time and machine time kept apartHow to model it
- Random seed and reproducibilityChecked against a known answer
- Simultaneous events handled in order
Capacity, resource pools and conflicts
- Capacity in units per timeSized by Build layoutChecked against a known answer
- Several jobs at once on one resourceSized by Build layout
- Resource classes with interchangeable instancesSized by Build layoutImproveChecked against a known answer
- Resource classes with differing properties
- A resource shared by several process stepsSized by Build layoutChecked against a known answer
- A resource shared by several productsSized by Build layout
- Mutually exclusive operating modes
- Allocation limited by compatibilitySized by Build layout
- Shared resource contention
- Deadlock detectionChecked against a known answer
Breakdowns and reliability
- No disturbances (baseline)Sized by Build layout
- MTBF and MTTR reliability modelSized by Build layoutChecked against a known answerHow to model it
- Fixed planned stops
- Random machine stopsChecked against a known answer
- Failures only while workingHow to model it
- Failures on calendar time
- Breakdowns that spoil the unit in workHow to model it
- Breakdowns where the cycle resumesHow to model it
- Failure that forces the work to restartHow to model it
- Failure dependencies
- Failure rate that rises with age (wear-out)How to model it
- A limited repair crew
Quality, deviations and rework
- Measurement and inspection as operationsSized by Build layoutHow to model it
Physical layout and space
- Station footprintSized by Build layout
- Walkways and distancesSized by Build layout
- Machine areas that never overlapSized by Build layout
- Layout changes that change the flowChecked against a known answer
- Distance to a shared resource
Automatic synthesis and sizing
- Number of stations from demandSized by Build layoutChecked against a known answer
- Number of parallel machinesSized by Build layoutChecked against a known answer
- Process route generated from the operationsSized by Build layoutChecked against a known answer
- Synthesis of hierarchical assemblySized by Build layout
- Synthesis for a mixed product mixSized by Build layout
- Synthesis with shared machinesSized by Build layout
- Synthesis with re-entrant flowsSized by Build layout
- Synthesis with alternative station typesSized by Build layoutImprove
- Synthesis with locked machinesSized by Build layout
- Synthesis with allowed duplicationSized by Build layout
- Synthesis limited by operatorsSized by Build layoutImprove
- Synthesis limited by floor areaSized by Build layout
- Synthesis that decides buffersSized by Build layoutImprove
- Synthesis with transport resourcesSized by Build layoutImprove
- Synthesis with availability and failuresSized by Build layout
- Synthesis to a service levelSized by Build layoutChecked against a known answer
- Synthesis to any target horizonSized by Build layoutChecked against a known answer
Optimisation and decision support
- Maximise throughputSized by Build layoutImprove
- Minimise machinesSized by Build layoutImproveChecked against a known answer
- Minimise operatorsSized by Build layoutImprove
- Minimise lead timeImprove
- Minimise investmentSized by Build layoutImprove
- Optimise buffer sizesSized by Build layoutImprove
- Optimise the transport fleetSized by Build layoutImprove
- Multi-objective Pareto frontImprove
- Optimisation aware of uncertaintyImproveChecked against a known answer
- Improve an existing manual layoutImprove
- Compare a candidate against a baselineImproveChecked against a known answer
Simulation engine and validity
- Event order at equal time stamps
- No resource booked twiceChecked against a known answer
- Material conservationChecked against a known answer
- Work and resource conservationChecked against a known answer
- Reproducible with a seedChecked against a known answer
- Variation between independent seedsChecked against a known answer
- Monte Carlo replicationsChecked against a known answer
- Confidence intervalChecked against a known answer
- Warm-up and transientsChecked against a known answer
- End conditions: time, units or finished orders
- Running until the last order is done
- Deadlock and livelockChecked against a known answer
- No negative times
- Numerical stability over long horizonsChecked against a known answer
- Infinite queues handled
- Insensitive baseline caseChecked against a known answer
- Little's law checkChecked against a known answer
- Static constraints validatedSized by Build layout
- Deterministic rule orderChecked against a known answer
- Sensitivity analysisChecked against a known answer
Measures and explanation of results
- Throughput of good unitsHow to model it
- Ideal and actual rateChecked against a known answer
- Cycle time and lead time
- Queue and waiting time per station
- Work in process over time
- Utilisation per resource
- OEE: availability, performance and qualityHow to model it
- Blocked and starved time
- Bottleneck identificationChecked against a known answer
- Bottleneck that moves over timeHow to model it
- Order lead time and punctuality
- Service level against deadlines
- First-pass yield
- Queue percentiles and extremes
- Timeline, Gantt and event log
- Resource cost and cost per good unit
- Warning for saturated or unstable queues
- Explanation of sizing decisionsSized by Build layout