PLANET of MIMICS

Use cases for DBM / CREEM

DBM/CREEM connects AI services and enterprise applications to OT through typed resources, callable capabilities, workflows, and trace records.

A project usually starts with namespace and asset contracts. Engineers then bind runtime context, test workflow behavior, and retain the resulting event history for diagnosis and later tuning.

Typical project conditions

Multiple plants or lines use different equipment generations, protocols, and system owners.
High-mix production requires frequent recipe, tooling, and routing changes.
High-volume operations need synchronized virtual commissioning and production trace records.

DBM artifact	Industrial responsibility	Runtime evidence
Resource Description (RD)	Identifies equipment, software services, people, materials, and their ownership or location context.	Resolved resource URI, type version, connector binding, and availability state.
Function Block (FB)	Defines a callable capability with typed ports, parameters, preconditions, completion criteria, and faults.	Invocation ID, bound inputs, output values, duration, result code, and exception path.
Workflow (WF)	Coordinates sequence, parallel work, retries, timeout, compensation, escalation, and operator decisions.	Workflow instance, active node, transition reason, retry count, and final disposition.
Data Table (DT)	Carries shared recipe, order, quality, environment, and policy context across behavior nodes.	Value source, unit, timestamp, quality code, version, and change history.
Behavior Tree / State Machine	Expresses execution policy and equipment lifecycle without mixing it with device protocol details.	Tick or transition trace, guard evaluation, state entry and exit, and recovery decision.

Cross-industry DBM application catalog

These scenarios use the same DBM primitives but bind them to different domain types, evidence requirements, and command authorities.

Industrial domain	Scenario	DBM application	Primary boundary
Pharmaceutical production	Electronic batch execution and deviation handling.	Bind approved master recipe versions to equipment capabilities; trace operator signatures, parameter changes, samples, and exception disposition.	DBM coordinates execution; validated systems retain record, signature, and release authority.
Water and wastewater	Pump scheduling, treatment mode changes, sampling, and storm response.	Coordinate reservoirs, pumps, chemical dosing, laboratory status, weather context, and energy constraints across sites.	PLC and protection logic enforce local process and equipment limits.
Mining and bulk materials	Crusher, conveyor, stockpile, and load-out coordination.	Model route availability, material grade, equipment capacity, blockage recovery, and maintenance windows as workflow context.	Machine protection and emergency stop remain local.
Data centers	Cooling optimization and maintenance transfer.	Sequence workload context, cooling topology, UPS state, redundancy class, and change approval before equipment isolation.	BMS, EPMS, UPS, and protection systems retain automatic control authority.
Ports and terminals	Berth, crane, yard vehicle, gate, and customs coordination.	Track vessel call, container identity, custody, hazardous classification, route, resource allocation, and delay reason.	Crane anti-collision and vehicle safety controllers execute motion limits.
Rail and transit	Depot inspection, cleaning, charging, and release-to-service.	Coordinate vehicle state, work orders, test results, staff qualifications, track allocation, and timetable deadlines.	Signaling and train protection are outside DBM command authority.
Renewable energy	Wind or solar fleet inspection and curtailment workflow.	Relate forecast, asset condition, grid request, access permit, technician dispatch, and restoration evidence.	Turbine, inverter, substation, and grid protection remain authoritative.
Buildings and campuses	Occupancy-aware HVAC, comfort complaints, and maintenance response.	Bind rooms, zones, meters, equipment, schedules, indoor quality, tickets, and energy policy to reusable workflows.	BMS controllers execute bounded setpoint and equipment commands.

Alarm system for maritime operations

An ocean-going cargo vessel distributes alarm handling across safety systems, machinery monitoring, navigation, power, communications, and operator stations. The alarm context must remain clear during a long voyage, heavy weather, crew handover, port operations, and remote support.

The structural view models a ship-wide alarm ontology. Machinery spaces, the bridge, cargo holds, deck equipment, power distribution, fire and flooding protection, communication equipment, and crew procedures become resources with typed signals, thresholds, escalation rules, and acknowledgement states.

Alarm scope	Typical signals	DBM/CREEM interpretation
Propulsion and machinery	Main engine temperature, lubrication pressure, vibration, fuel system status.	Equipment resources expose behavior states such as normal, warning, derated, trip, bypass, and recovery.
Electrical and power	Generator load, switchboard status, breaker trip, UPS health, blackout indicators.	Power topology becomes a dependency graph for alarm propagation and root-cause ranking.
Navigation and voyage safety	Bridge alerts, heading deviation, AIS/GPS status, weather-window risk, restricted-area warning.	Voyage context binds alarms to route phase, sea condition, watch duty, and maneuvering mode.
Cargo and environmental protection	Hold temperature, reefer status, bilge level, fire detection, emission or ballast constraints.	Alarm semantics connect cargo state, compliance boundary, equipment action, and evidence trace.

Embedded sensors and PLC or RTU endpoints publish raw events. The shipboard alarm server assigns lifecycle state and vessel context, while DBM models the response workflow. Shore services receive the resulting alarm record with its source, state, dependencies, and response history.

The event lifecycle records detection, classification, vessel context, priority, notification, acknowledgement, response, recovery verification, and closure.
Correlation rules compare repeated alarms with equipment state, voyage phase, weather, maintenance history, and crew action.
An AI agent receives the source asset, affected subsystem, severity, dependency path, approved checklist, evidence, and confidence score.
Fleet teams can compare alarm frequency, response time, recurrence, and unresolved causes without discarding vessel-specific logic.

DBM/CREEM links each alarm to vessel structure, operating state, and response procedures. The runtime records the workflow and its decisions alongside the original signals.

Cross-domain orchestration

One workflow coordinates ERP and MES data with PLC signals, robot motion, machine vision, AGV dispatch, and database updates. The workflow records when each system accepts work, returns a result, or reports a fault.

Namespace bindings keep protocol and device details outside the orchestration model. Engineers can map a compatible simulator or device to the same capability contract.

Order trigger from ERP/MES.
Recipe + device parameter binding.
OT execution with camera/robot/PLC coordination.
Result write-back to historian and business systems.

High-mix production for automotive lines

For small-batch and frequent changeover scenarios, users can reconfigure behavior trees and parameter mappings without rewriting large software stacks.

Runtime recipe switching through parameterized FB ports.
Workflow branch reuse across multiple product variants.
Commissioning tests scoped to the branches and parameters changed for each variant.

Batch recipe orchestration for chemical and food production

A batch workflow coordinates material verification, dosing, heating, agitation, hold time, sampling, discharge, and cleaning. Equipment capability stays separate from the product recipe, so the scheduler can allocate the same recipe to any qualified unit.

Map unit procedures to workflows, phases to FB capabilities, and recipe values to a versioned Data Table.
Check material status, equipment state, interlocks, utility availability, operator authority, and quality release before execution.
Define hold, controlled stop, alternate equipment, material quarantine, and supervised resume paths.
Record consumed lots, actual parameters, phase transitions, samples, deviations, approvals, and cleaning status.

Flexible robot cell recovery

A robot cell must recover from failed picks, vision uncertainty, missing parts, tool faults, and safety interruptions without losing order or workpiece context. DBM models each recovery path as an explicit branch rather than burying it in device-specific code.

Condition	Behavior response
Low vision confidence	Recapture with adjusted exposure, request alternate model, then route to operator review.
Pick verification failed	Return to safe pose, inspect gripper state, retry within policy, or reject the part.
Safety stop	Freeze workflow context, verify reset authority, revalidate cell state, and resume from an approved node.
Tool unavailable	Evaluate compatible tools or cells and rebind the capability when the process contract permits it.

Brownfield line modernization

Existing lines often combine PLC generations, proprietary robot interfaces, manual stations, spreadsheets, and MES transactions. DBM introduces a typed behavior layer above these endpoints while leaving validated machine control in place.

Wrap existing commands and signals as FB ports with explicit units, quality, timeout, and ownership.
Map legacy tags to stable resource identities instead of exposing raw addresses to workflows.
Run new and old sequences in shadow mode before enabling command authority.
Record connector version and fallback route for each migrated behavior.

Physical and virtual consistency

The same DBM workflow can drive both virtual assets and physical equipment by binding through namespace and runtime context.

Engineers can test sequence and timing against a simulator, run the approved model with live signals, and replay recorded context after execution.

Binding target	Runtime effect
Digital twin	Validate sequence and timing before release.
Physical cell	Execute the same model with live signals.
Replay service	Reconstruct history for analysis and training.

Intralogistics fleet coordination

AMRs, AGVs, conveyors, lifts, doors, and manual forklifts share routes and transfer points. A DBM workflow coordinates transport intent while fleet managers and safety controllers retain responsibility for motion execution.

Carry source, destination, load class, priority, due time, and custody state with the transport order.
Select a qualified transport resource by payload, geometry, zone access, charge state, and availability.
Coordinate doors, lifts, conveyors, machine buffers, and receiving confirmation as separate behaviors.
Reroute, wait, reassign, return the load, or escalate when a route or destination becomes unavailable.

Industrial energy and microgrid dispatch

A site workflow can coordinate production loads with photovoltaic generation, battery storage, generators, charging infrastructure, tariff windows, and demand limits. Electrical protection and certified controllers remain the command authority.

DBM context	Example decision
Production schedule	Move discretionary heating, charging, or compressed-air work within an allowed time window.
Power envelope	Apply a site demand cap and preserve reserve for safety-critical loads.
Asset state	Exclude batteries or generators that are unavailable, constrained, or under maintenance.
Evidence	Record forecast input, selected policy, command request, measured response, and operator override.

Condition-based maintenance workflow

Condition monitoring becomes actionable when the workflow links an observation to asset identity, operating mode, failure hypotheses, maintenance history, production constraints, and approval rules.

Qualify vibration, temperature, oil, acoustic, or electrical observations before using them in a decision.
Run diagnostic branches against the current operating state and known maintenance conditions.
Create an inspection or work-order proposal with evidence, confidence, urgency, and required skill.
Coordinate derating or planned shutdown only through approved operations and maintenance workflows.

Worked example: screw fastening cell

A single composite workflow can coordinate robot homing, camera capture, vision detection, screw pick-and-insert, quality test, retry policies, and operator escalation.

The industrial behavior-tree example in the deep dive uses this pattern to place AI, IT, and OT calls in one workflow.

Engineers can map the sequence to a simulator, a FAT setup, or a production line. Each target supplies its own connector binding while preserving the behavior contract.

Parallel homing and safety checks.
Vision capture, preprocessing, and registration.
Insert/fasten sequence with quality gate and retry path.
Operator escalation branch and recycle strategy handling.

Safety and command authority

DBM can coordinate industrial intent, but it should not replace certified protection or embedded equipment control. Each Function Block contract must state where command authority begins and ends.

The safety PLC or SIS owns trips, permissives, emergency shutdown, and safety integrity functions.
The machine controller owns servo loops, interlocks, motion limits, and equipment-level sequence integrity.
The DBM runtime requests capabilities, coordinates resources, applies workflow policy, and records decisions.
An AI service can propose classifications, plans, or parameter candidates within an approved contract. It cannot bypass authorization.
The operator retains the approval, override, hold, and recovery duties assigned by site procedure.

Regulated batch and genealogy

Food, pharmaceutical, battery, and specialty-chemical production require more than sequence execution. The workflow must preserve which material, method, equipment, software version, and approval produced each result.

Resolve material lots and equipment status before phase release.
Capture actual values beside approved targets and tolerances.
Link deviations to affected workflow nodes, lots, samples, and disposition decisions.
Prevent workflow continuation until required quality or operator gates are satisfied.
Export a replayable genealogy package. Treat DBM as the system of record only when the project has validated it for that role.

Remote operations and fleet rollout

A central team can publish versioned behavior templates to vessels, plants, warehouses, or energy sites while each location keeps its own resource bindings and operating limits.

Control	Required record
Template release	Model version, change reason, test evidence, approver, and compatible resource types.
Site binding	Local URI, connector version, capability limits, units, and authority owner.
Deployment	Target instances, rollout window, health checks, rollback package, and status.
Runtime comparison	Common behavior metrics with site-specific operating context and exclusions.

Deployment measures

Measure how many RD and FB definitions project teams reuse without project-specific changes.
Count custom connector mappings and workflow branches required for each deployment.
Track the engineering and test effort needed to qualify a new device or product variant.
Check whether incident reviews can reconstruct state, inputs, decisions, and operator actions from the trace.

Project adoption checklist

Define namespace and type-governance rules first.
Prepare reusable RD/FB libraries for target domains.
Build connector mappings for key IT/OT endpoints.
Start from one pilot cell, then scale by workflow reuse.

Adoption Stage	Acceptance check
Pilot	Validate orchestration semantics and connector reliability.
Scale-out	Reuse RD/FB/WF assets across additional lines or cells.
Optimization	Use runtime event history to improve throughput and quality.

OT behavior traceability

Global data tables and temporal event streams capture behavior-state transitions for diagnostics, quality tracing, and lifecycle analytics.

Regulated projects use these records for root-cause analysis, audits, and incident closure across operations, engineering, and quality teams.

Event timeline replay by namespace and workflow instance.
State transition tracking for each behavior node.
Correlation of quality results with execution context and process parameters.

Use-case verification strategy

Engineers test a DBM application at the contract, workflow, integration, and operational levels before enabling command authority.

Verification level	Checks
Type and port contract	Data type, unit, range, null handling, quality code, timeout, and backward compatibility.
Behavior node	Precondition, success, failure, cancellation, idempotency, retry safety, and resource cleanup.
Workflow model	Branch coverage, parallel synchronization, deadlock, timeout, compensation, hold, resume, and escalation.
Connector integration	Protocol mapping, stale data, duplicate events, reconnect behavior, command acknowledgment, and clock alignment.
Virtual commissioning	Nominal cycle, boundary conditions, injected faults, performance envelope, and operator interaction.
Site acceptance	Authority matrix, safety interface, rollback, audit record, support handover, and production release criteria.

AI engineering assistant

DBM gives an AI assistant typed resources, capability contracts, workflow structure, and execution evidence. The assistant can use this context for engineering and diagnosis without receiving unrestricted control.

Draft RD, FB, Data Table, and workflow definitions from approved domain templates.
Identify incompatible ports, missing units, unreachable branches, absent timeout handling, and unresolved resource references.
Derive nominal, boundary, and fault-injection cases from behavior contracts.
Correlate event traces with asset state, workflow transitions, maintenance history, and current operating context.
Propose parameter or workflow changes with expected effect, evidence, confidence, and rollback conditions.
Send proposals through deterministic validation and human or policy approval before execution.