TOKNITH
Operational modeling and execution platform for complex physical systems.

The operation is one system.

One model. One clock. Reality and rehearsal.

Discrete work happens at a time.

Continuous processes change through time.

Agents move, interact, and respond.

TOKNITH brings them together in one operational model.

Model it. Run it. Branch it. Connect it to reality.

EVENTSCONTINUOUSAGENTSONE CLOCKOP1AGV107:1007:1507:2007:2507:0407:04 repair accepted · CF0001 · shift_leadF2H0 to F2H3 cleared by OP1trucks leave on their slots24 bottles/min48tank T1 0.68 m0.74 mOP1 walks to and repairs 8 heads, 2 min eachhomeidlebottles per minutetank T1 levelhead repairwalkhead clearedYogurt line demo shift, Advisor on · 07:04 to 07:25 · every mark is a value the model publishedNOW

Events · continuous processes · agents · one clock

A warehouse. A hospital. A factory. A complex facility.

Different operations. Same fundamental problem:

Physical systems and operational work have to interact in time.

A warehouse

An order moves through a warehouse while conveyors continue running and workers move through the facility.

A hospital

A patient moves through a hospital while beds, rooms, elevators, staff, equipment, and treatment processes change continuously.

A factory

A part moves through a factory while machines operate, buffers fill, operators move between work centers, and failures change what happens next.

The operation is not just a workflow. It is a system.

TOKNITH models the system as one.

Discrete + Continuous + Agent-based

Three ways an operation behaves.

Real operations are not purely discrete or purely continuous. They contain different kinds of behavior that interact constantly.

Discrete

Events happen.

An order releases. A job completes. A truck arrives. A machine fails. A patient is assigned a bed. A wave closes. A shipment leaves.

Continuous

Processes evolve.

A conveyor moves. A tank fills. A temperature changes. A battery charges. A buffer accumulates. A vehicle travels.

Agent-based

Agents act.

A worker moves through a facility. A forklift chooses its next task. An AMR travels between locations. A patient moves through care.

The three drawings above are values the model published: the yogurt line demo shift and the warehouse pack.

TOKNITH resolves all three within one model and one clock.

These are not three separate simulations. They are three forms of behavior within one operational model. Agents here means agent-based modeling, not AI agents.

TOKNITH · תָּכְנִית

The measured pattern of the operation.

The name TOKNITH comes from the Hebrew toknith, associated with a measured plan, pattern, proportion, or complete design.

TOKNITH is not simply a picture of an operation. It is a model of how the operation is supposed to behave. The model can be run, examined, changed, branched, compared, and ultimately connected to the real operation.

The model is the authority.

  • Physical system
  • Operational system
  • Time
  • Behavior
  • State
THE MODEL
  • Run
  • Branch
  • Compare
  • Connect
Visualization

The picture is not allowed to lie.

Visualization is not a separate presentation layer that invents what the operation looks like. It is a representation of model state.

MODEL STATEyogurt line · Advisor offclock07:10T1.level_m0.711B2.level_m0.137C1.speed_m_s0.4F1H0–F1H3.fault1 · 1 · 1 · 1F2H0–F2H3.fault1 · 1 · 1 · 1LP1.parts_on1LP1.next_truck07:12PAL1.busy0OP1.stateidleconductor.batch · phase3 · RUNconductor.orders_open9fine_duty0.0DERIVESVISUALIZATION07:101 m0T10.711 mB20.137 mF1F2heads 0 to 3 faulted in both banksC1 · 0.4 m/sPAL1 · idleLP1 · 1 pallet · next truck 07:12OP1 · idlebatch 3 · RUN · 9 orders openfine duty 0.0every mark on the right is a value on the left
  • If the model says the liquid rose, the picture shows it rising.
  • If the model says a worker is blocked, the picture shows the worker blocked.
  • If the model says a conveyor backed up, the accumulation exists.
  • If reality disagrees with the model, that disagreement belongs to the model.
Range · the edition without feeds

Rehearse before you change.

Range is the twin without the feeds. The model, the schedules, the what-if branches, and the studies. Build the operation. Run it. Change it. Branch it. Compare alternatives. Test failures. See what happens.

Range explores the operation without making the real operation pay for the experiment.

Range answers: What could happen?

  1. Model
  2. Branch
  3. Run
  4. Compare
BASELINEbranchBRANCH+ customer · + AMRs · − 1 shiftCOMPAREthroughputcongestionlabor hrsbaselinebranch · not live until a person acceptsILLUSTRATION
  1. Expected
  2. Observed
  3. Residual
  4. Understand
TIMECYCLE TIMERESIDUALcycle time · +12 %expected (model)observed (floor)residual · investigated, not hiddenILLUSTRATION
Campus · Range plus the connection

Connect the model to reality.

Campus is everything in Range, plus the connection. The same model, the same objects, the same process definitions, the same clock, the same physics.

Now the operation provides observations, and TOKNITH compares what should be happening with what is happening.

  • The real-time view. Expected state beside observed state, on one clock.
  • Divergence checks by object. Error named on the object that owns it.
  • The Advisor, as an add-on. Proposals a named person accepts or rejects.

Campus answers: What is happening?

Advisor · available on Campus

The Advisor proposes. It never applies.

Candidate actions are run as branches of the same model and scored against doing nothing. Only the ones that beat it are proposed. A named person accepts, the commit goes through your own systems, and rejecting changes nothing.

Range → Campus

From rehearsal to operation.

The model used to evaluate the change becomes the model used to understand the resulting operation. TOKNITH does not throw the model away when the operation goes live.

  1. Model
  2. Range
  3. Branch
  4. Compare
  5. Accept
  6. Campus

A rehearsal does not silently become live. A person accepts the change.

Accepting is the boundary. It separates what is possible, what is expected, what has been approved, and what is actual.

Not just simulation.

TOKNITH uses simulation as a capability, not as the category.

Traditional approachTOKNITH
DES models eventsEvents + continuous processes + agents
Physics simulation models physical behaviorPhysical + operational behavior
Digital twin represents a facilityExecutable operational model
Workflow systems manage workModel the physical consequences of work
Visualization shows stateVisualization derives from state
Scenario tools create isolated simulationsBranches derive from the operational model
Live systems report what happenedModel compares expected and observed behavior

The difference is not another feature. It is the model.

The moat

The moat is the model, and the fact that everything else is required to agree with it.

The same state drives the picture. The same clock advances the operation. The same model receives observations from the floor, records disagreement, and can be branched to explore another future.

Read the architecture

One clock. One model. Many futures. One reality.

Model the operation. Rehearse the change. Connect it to reality.

Request a Demo