This page defines the core terminology used in the TPCA / PCN state-transition pre-control framework.

The terminology is intended for public technical explanations, application cases, engineering discussions, and future technical documentation.

The core proposition of TPCA / PCN is:

A clearly defined Target State Entry should be treated as an engineering object that can be independently designed, evaluated, controlled, and recorded.

Core Terminology

TermMeaningBasic Role
TPCATransition Pre-Control ArchitectureOverall pre-control architecture organized around a clearly defined Target State Entry
PCNPre-Control NodeEngineering node placed before a specific Target State Entry
Current StateCurrent stateThe state, stage, or path position currently occupied by the system
Target StateTarget stateThe next state, execution path, or physical execution stage that the system is intended to enter
Target State EntryTarget-state entry pointThe engineering point at which pre-control is performed before entering the Target State
CConditionDetermines whether the factual prerequisites required for the Target State are satisfied
AAuthorityDetermines whether the system is permitted to enter the Target State
EExecution ChainDetermines whether the execution chain can continue after entry into the Target State
SStructureEvaluates whether required signals, interfaces, mappings, authority sources, and execution-chain boundaries are structurally established
DDynamicsEvaluates whether relevant states remain dynamically valid, synchronized, stable, current, and effective
BBoundaryEvaluates whether relevant states remain within predefined control boundaries
CAE-SDBStructured evaluation logicCombines C / A / E state roles with S / D / B evaluation properties
CAE-SDB ResultStructured evaluation resultOne or more structured results produced by a pre-control evaluation
TTime informationTime-related information retained with states and evaluation results
ArbitrationControl arbitrationResolves priorities among evaluation results and predefined control constraints
Multipath ControlMulti-path control outputDetermines the controlled path following Arbitration
PCN TraceState-transition decision traceRecords the inputs, evaluations, time information, control output, and execution result of a Target State Entry
PCN RuntimePCN runtimeRuntime layer that performs online evaluation, arbitration, control output, and trace generation
PCN NetworkPCN networkA pre-control structure formed by multiple PCNs connected through state-transition and dependency relationships

The basic engineering relationship is:

Current State
→ Target State
→ Target State Entry / PCN
→ Multi-source State Signals
→ C / A / E Mapping
→ S / D / B Evaluation
→ CAE-SDB Result + T
→ Arbitration
→ Multipath Control
→ PCN Trace

PCN Runtime executes this process in an operating system.

Multiple PCNs can be connected according to their Target State Entries and state-transition relationships to form a PCN Network.

Current State, Target State, and Target State Entry

Current State is the state, stage, or path position occupied by the system before a state transition.

Target State is the next state, target execution path, or target physical execution stage that the system is intended to enter.

Target State Entry is the engineering point immediately before the system moves from the Current State into the Target State, where pre-control evaluation is performed.

TPCA / PCN focuses on the engineering decision made before the system actually enters the Target State.

Typical transitions include:

  • Waiting Stage → Pick Stage
  • Placement Complete → Pressing Stage
  • Inspection Waiting → Inspection Execution
  • Task Exists → Task Execution
  • AGV Arrival → Station Acceptance
  • Current Request → Target Invocation Path

A Target State is not limited to the next state in the normal production flow.

Rework, return flow, fallback, exception routing, re-entry, alternative execution paths, and other states that require a new entry decision may also become Target States.

From the runtime perspective of TPCA, these are represented as:

Current State → New Target State

Even when a later Target State has the same or similar engineering content as a previous state, it represents a new state instance in actual operation.

The Current State, Target State, and Target State Entry must therefore be clearly defined. Without them, a PCN does not have a definite object for pre-control evaluation.

Pre-Control

Pre-Control refers to the structured evaluation of relevant multi-source states before a system enters a Target State, target execution path, or target physical execution stage.

Based on the complete evaluation result, the system determines whether the transition may proceed and which controlled path should be taken.

Pre-Control can include evaluation of:

  • whether required conditions are satisfied;
  • whether required authority is established;
  • whether the execution chain can continue;
  • whether the required engineering structure is complete;
  • whether the current states remain dynamically valid;
  • whether predefined control boundaries have been reached;
  • how multiple evaluation results should be arbitrated;
  • which Target State or controlled execution path should follow.

Pre-Control occurs before entry into the Target State.

Its engineering position is therefore different from post-failure diagnosis or simple alarm classification.

TPCA

Transition Pre-Control Architecture

TPCA is the overall architecture for pre-control around a clearly defined Target State Entry.

It organizes states that may otherwise be distributed across equipment, control programs, MES / WCS, safety systems, manual confirmation processes, and other engineering interfaces around the same Target State Entry.

The basic engineering chain is:

Current State
→ Target State
→ Target State Entry / PCN
→ C / A / E Mapping
→ S / D / B Evaluation
→ CAE-SDB Result + T
→ Arbitration
→ Multipath Control
→ PCN Trace
PCN执行示意动图
1 / 8
PCN执行示意动图多源设备与系统信号进入 PCN，在 PCN 内部映射为 C、A、E 三个变量域， 再通过 S、D、B 三类判定性质形成九个 CAE-SDB 判定坐标。 一个或多个判定坐标被触发后进入 Arbitration，并选择相应执行路径。 示例先触发 E-D，进入 Wait / Refresh；状态刷新后重新判定，进入正常执行链并到达 Target State。多源设备 / 系统信号PCN：CAE-SDB 判定执行路径 / Target StateVision_ResultWork_Order / TaskSafety_PermitQuality_ReleaseRobot_ReadyDownstream_AvailablePCNPre-Control NodeC / A / E MappingCConditionAAuthorityEExecution ChainVision / Work OrderSafety / QualityRobot / DownstreamCAE × SDBDomainSDBCAEC-SC-DC-BA-SA-DA-BE-SE-DE-BActive result set: noneCAE-SDB ResultE-DDownstream state not refreshedArbitrationWait / RefreshWait / RefreshRecheckManual ConfirmProhibitNormal Execution ChainTarget execution pathTarget StateNot EnteredPCN Trace #0001271st: E-D → Wait / RefreshState refreshedDownstream_Available = valid
多源信号准备进入 PCN

示意：PCN 将多源状态映射到 C / A / E，并从 S / D / B 三类性质形成 9 个判定坐标。 一个或多个判定坐标被触发后，经 Arbitration 选择相应执行路径；正常执行链成立时进入 Target State。

TPCA does not replace PLCs, state machines, SFCs, interlocks, safety control, MES, or WCS.

When multiple existing mechanisms jointly determine whether a state transition can occur, TPCA organizes their relevant states, pre-control evaluation, control result, and decision trace around one explicit Target State Entry.

TPCA also retains time-related information associated with states and evaluations.

Real system transitions proceed forward in time. Even when a later state has the same engineering content as an earlier state, it represents a new state instance in actual operation.

PCN

Pre-Control Node

A PCN is an engineering node placed before one clearly defined Target State Entry.

One PCN corresponds to one explicit Target State Entry and can perform:

  • confirmation of the Current State and Target State;
  • acquisition and organization of multi-source states;
  • C / A / E Mapping;
  • S / D / B Evaluation;
  • generation of CAE-SDB Result;
  • retention of time information T;
  • Arbitration;
  • Multipath Control output;
  • generation of PCN Trace.

The implementation form of a PCN is not fixed.

In an automated execution unit, a PCN may be implemented as a PLC function block, an industrial edge-control module, or a software evaluation node.

In MES / WCS, AGV / AMR fleet control, manufacturing DX, or digital systems, a PCN may be placed before task execution, resource release, station acceptance, state recovery, or entry into a target invocation path.

PCN Runtime

PCN Runtime is the runtime layer that performs online state processing, evaluation, control output, and trace generation for a PCN.

For a specific Target State Entry, PCN Runtime can perform:

  • state acquisition and organization;
  • C / A / E Mapping;
  • S / D / B Evaluation;
  • generation of CAE-SDB Result;
  • retention of time information;
  • Arbitration;
  • Multipath Control output;
  • PCN Trace generation.

PCN Runtime describes a runtime role rather than a fixed implementation platform.

It may be implemented in software, PLCs, industrial edge controllers, MES / WCS plug-ins, or other suitable systems.

Detailed Runtime architecture, configuration structures, Rule Packs, complete Arbitration logic, control priorities, key parameters, interfaces, and deployment implementation are outside the general public scope of this website.

PCN Network

A PCN Network is a pre-control structure formed by multiple PCNs connected according to actual state-transition relationships and the necessary authority, resource, and execution dependencies between them.

A single PCN corresponds to one explicit Target State Entry.

Multiple PCNs may be distributed across equipment units, production lines, MES / WCS, AGV / AMR fleet-control layers, manufacturing DX systems, or digital systems.

A PCN Network describes engineering relationships between Target State Entries.

It is not defined simply by the physical connections between devices.

For example:

State A
→ PCN-1
→ State B
→ PCN-2
→ State C

When a system enters rework, return flow, exception routing, re-entry, or another execution path, a new Target State Entry and the corresponding PCN may also be formed.

State-type relationships within a PCN Network may form cycles, while actual state instances continue to be generated forward in time.

Such cycles describe relationships among state types or transition paths, not a reversal of runtime history. Each actual state instance occurs at a new point in time and remains uniquely traceable.

A PCN Network can be used to observe relationships involving:

  • state progression;
  • authority dependencies;
  • resource dependencies;
  • execution-chain continuity;
  • state write-back;
  • decision traces.

Multi-source State Signals

Multi-source state signals are input information directly relevant to the current Target State Entry.

Signal sources may include:

  • PLCs;
  • robot controllers;
  • vision systems;
  • safety systems;
  • HMI / SCADA;
  • MES;
  • WCS;
  • AGV / AMR fleet-control systems;
  • field sensors;
  • supervisory systems;
  • manual confirmation;
  • digital-system interfaces.

A PCN maps relevant states into C / A / E according to the engineering role that each state plays in the current Target State Entry.

The mapping is not determined simply by the source system.

The same signal may play different roles at different Target State Entries.

Multi-source state information must therefore be interpreted together with the Current State, Target State, and Target State Entry.

Depending on the system, state information may also include:

  • timestamps;
  • update times;
  • state versions;
  • sequence information;
  • lot numbers;
  • object IDs;
  • event IDs;
  • other traceability information related to the current transition.

CAE-SDB

CAE-SDB is the structured evaluation logic used within a PCN.

Its basic structure is:

{C, A, E} × {S, D, B}

where:

  • C / A / E represent state-variable roles;
  • S / D / B represent evaluation properties applied to those states.

Together they define a 3 × 3 available evaluation space:

SDB
CC-SC-DC-B
AA-SA-DA-B
EE-SE-DE-B

A specific pre-control evaluation may produce one or more CAE-SDB Results, together with the relevant time information T.

A CAE-SDB Result is produced only when the corresponding evaluation has actually been executed and has produced a result.

When multiple results exist at the same Target State Entry, the relationship is:

CAE-SDB Result + T
→ Arbitration
→ Multipath Control

CAE-SDB expands compressed operational states such as:

false
NG
not ready
waiting
pending
blocked

into structured engineering results that identify both:

  • the state-variable role involved; and
  • the type of evaluation that produced the result.

C: Condition

Condition

C represents whether the factual conditions required before entering the Target State are satisfied.

Typical examples include:

  • whether a workpiece exists;
  • whether position requirements are satisfied;
  • whether recognition results are valid;
  • whether a task exists;
  • whether required parameters are complete;
  • whether the preceding stage is complete;
  • whether product or lot information is correct;
  • whether the object can be correctly associated with the corresponding result.

C focuses on the factual prerequisites required for the Target State Entry.

A: Authority

Authority

A represents whether the system is permitted to enter the Target State.

Typical examples include:

  • safety permission;
  • supervisory-system permission;
  • area permission;
  • manual confirmation;
  • access rights or authorization;
  • resource locks;
  • permission from another device or system.

A critical Authority state can constitute an independent necessary constraint.

If a critical permission such as safety authorization, access authorization, resource ownership, or area permission is not established, the system must not enter the Target State even when C and E are otherwise satisfied.

A therefore answers:

Is the system allowed to enter this Target State?

E: Execution Chain

Execution Chain

E represents whether the execution chain associated with the current Target State can continue after the system enters that state.

E is not equivalent to the Ready state of an individual device.

Depending on the current Target State Entry, the execution chain may include:

  • the primary equipment;
  • the end effector;
  • the required motion or execution path;
  • downstream acceptance;
  • counterpart-device acceptance;
  • result transmission;
  • result write-back;
  • other links required for the current Target State to continue successfully.

For example, a robot may report:

Robot Ready = TRUE

but this alone does not demonstrate that the complete execution chain for the current Target State is available.

E must always be evaluated in relation to the current Target State Entry.

Alternative, fallback, return-flow, or exception-routing paths may become separate candidate Target States or execution paths.

The availability of such an alternative path does not, by itself, establish E for the current Target State and does not justify an Allow decision for the current Target State Entry.

E therefore answers:

Can the execution chain continue after entering this Target State?

S: Structure

Structure

S is an evaluation property applied to C / A / E states.

S determines whether the engineering structure required for the Target State Entry has been defined, connected, and made observable.

Typical structural issues include:

  • required signals are undefined;
  • required interfaces are not connected;
  • mapping relationships have not been established;
  • the source of Authority is unclear;
  • execution-chain boundaries are undefined;
  • required states cannot be observed;
  • relationships between objects and states have not been established.

S therefore evaluates whether the engineering structure required for the transition decision is complete.

D: Dynamics

Dynamics

D is an evaluation property applied to C / A / E states.

D determines whether relevant runtime states remain valid, synchronized, stable, current, and reliable.

Typical dynamic issues include:

  • timeout;
  • no refresh;
  • stale state;
  • oscillation;
  • conflict;
  • delay;
  • lack of synchronization;
  • version mismatch;
  • abnormal sequence relationship;
  • low confidence;
  • permission revocation;
  • a state currently undergoing transition.

A state value being logically true does not necessarily mean that it remains valid for the current Target State Entry.

For example:

Downstream Acceptable = TRUE

If the state has not been refreshed for an extended period, or was generated before a later downstream state change, the current value alone cannot establish that the execution chain remains valid.

D can therefore use information such as:

  • timestamps;
  • update times;
  • sequence information;
  • versions;
  • object associations.

D answers:

Can this state still be used as valid evidence for the current Target State Entry?

B: Boundary

Boundary

B is an evaluation property applied to C / A / E states.

B determines whether a relevant state remains within a predefined control boundary.

Engineering parameters used for Boundary evaluation may include:

  • allowable ranges;
  • time windows;
  • retry limits;
  • confidence ranges;
  • position-deviation limits;
  • manual-confirmation validity periods;
  • buffer capacity;
  • degraded-operation boundaries;
  • prohibition thresholds;
  • other engineering limits associated with the Target State Entry.

B itself does not determine the final control action.

The Boundary result is passed to Arbitration together with other CAE-SDB Results.

CAE-SDB Result

A CAE-SDB Result is a structured result produced after applying S / D / B evaluation to C / A / E states.

For example:

C-D

indicates a Dynamics-related problem within the Condition domain.

E-S

indicates a Structure-related problem within the Execution Chain domain.

A single Target State Entry may produce multiple results, for example:

C-D + E-D

which indicates dynamic-validity problems in both the Condition and Execution Chain domains.

A CAE-SDB Result is created only when the corresponding S / D / B evaluation has actually been executed and produced a result.

Relevant time information T is retained together with the result.

CAE-SDB Result identifies:

  • which C / A / E state role is involved; and
  • which S / D / B evaluation property produced the result.

The result is then passed to Arbitration.

Time Information T

T represents time-related information retained together with the state and CAE-SDB Result.

Depending on the system, T may include:

  • Timestamp;
  • update time;
  • event time;
  • sequence time;
  • lot-related time;
  • other information that establishes temporal relationships between states.

T is used to:

  • identify the temporal position of the current state and evaluation;
  • determine ordering relationships among states;
  • support Dynamics evaluation;
  • form the PCN Trace.

From the runtime perspective of TPCA, even when a later state has the same engineering content as a previous state, it represents a new state instance when it occurs at a different point in time.

Arbitration

Arbitration processes the priority relationships among one or more CAE-SDB Results, critical Authority states, and predefined control constraints.

When multiple evaluation results exist at the same Target State Entry, the final control direction cannot be determined by a single CAE-SDB coordinate alone.

The engineering position of Arbitration is:

CAE-SDB Result + T
→ Arbitration
→ Multipath Control

Relevant time information T is retained as part of the Arbitration context, allowing control rules to distinguish states with different timing, duration, freshness, or boundary conditions.

This website publicly explains the engineering role and position of Arbitration.

Detailed control priorities, short-circuit mechanisms, rule configuration, and complete arbitration implementation are outside the general public scope.

Multipath Control

Multipath Control is the final engineering control output produced after the complete CAE-SDB Result has been processed through Arbitration.

Representative controlled paths may include:

  • Allow;
  • Wait;
  • Re-identify;
  • Re-sample;
  • Reposition;
  • Return flow;
  • Fallback;
  • Exception routing;
  • Downstream coordination;
  • Resource release;
  • Degraded execution;
  • Alternative path;
  • Prohibit;
  • Safety lock;
  • Manual confirmation;
  • Exception isolation;
  • Enhanced recording.

Different control paths correspond to different subsequent engineering actions.

From the TPCA state-transition perspective, Multipath Control determines what controlled state or execution path should follow the current evaluation.

The basic relationship is:

CAE-SDB Result + T
→ Arbitration
→ Multipath Control

Authority (A) is an input state used in the pre-control evaluation, whereas Allow is a final control output produced only after Arbitration. Therefore, A being satisfied does not by itself result in an Allow decision.

All Allow, Prohibit, and other Multipath Control outputs are bound to the current explicit Target State Entry.

An alternative, fallback, or return-flow path may be selected as a Multipath Control candidate.

However, the availability of such an alternative path does not mean that E for the current Target State has been established, and it does not justify an Allow decision for the current Target State Entry.

PCN Trace

PCN Trace is the structured record of a state-transition pre-control evaluation and its resulting control outcome.

Its record unit corresponds to one Target State Entry.

Representative contents may include:

  • Current State;
  • Target State;
  • multi-source states used in the evaluation;
  • time information T;
  • versions, sequences, or object information associated with the states;
  • C / A / E Mapping;
  • S / D / B Evaluation;
  • CAE-SDB Result;
  • Arbitration Result;
  • Multipath Control;
  • execution result;
  • Trace ID.

PCN Trace can support:

  • shop-floor review;
  • problem tracing;
  • state-transition design review;
  • project handover;
  • long-term comparison;
  • engineering improvement.

Its basic record relationship is:

Which Target State was being entered?
→ Which states were used as evidence?
→ Which structured results were produced?
→ How were those results arbitrated?
→ Which control path was selected?
→ Which state or path was actually reached?

Time information allows PCN Trace to distinguish state instances that have the same engineering content but occur at different points in time.

The detailed PCN Trace data structure is outside the general public scope of this website.

Relationship Between the Core Concepts

The relationship between the main TPCA / PCN concepts can be represented as:

TPCA
│
├─ PCN
│   │
│   ├─ PCN Runtime
│   │   │
│   │   ├─ C / A / E Mapping
│   │   ├─ S / D / B Evaluation
│   │   ├─ CAE-SDB Result + T
│   │   ├─ Arbitration
│   │   ├─ Multipath Control
│   │   └─ PCN Trace
│   │
│   └─ corresponds to one explicit Target State Entry
│
└─ Multiple PCNs
    ↓
   PCN Network

In summary:

  • TPCA: the overall architecture;
  • PCN: the engineering node associated with a specific Target State Entry;
  • PCN Runtime: the runtime layer that performs online pre-control evaluation and control;
  • C / A / E: state-variable roles;
  • S / D / B: evaluation properties;
  • CAE-SDB Result: structured evaluation result;
  • T: time information retained with states and results;
  • Arbitration: control arbitration;
  • Multipath Control: controlled output path;
  • PCN Trace: the trace of one Target State Entry evaluation;
  • PCN Network: a state-transition pre-control network formed by multiple PCNs.

Further Reading

Additional English application cases and technical notes will be added progressively.

This page is part of the public technical documentation of the TPCA / PCN state-transition pre-control framework.