Digital Twin Measurement Ontology
AI Measurement Semantics
Authors: MArk Anthony Higgins - Sky Recursive Ai
Revised Date: 16 September 2026
Audio Overview
# Digital Twin Measurement Ontology — Flow Algebra
URL: https://kirethresearch.org/booksai-core/ontologies/mathematical-ontologies/measurement-units-ontologies/ai-measurement-semantics/digital-twin-measurement-ontology.html
FLOW DigitalTwinMeasurementOntology {
DOMAIN:
Mathematical Ontologies
-> Measurement & Units Ontologies
-> AI Measurement Semantics
-> Digital Twin Measurement Ontology
PURPOSE:
Represent measurement semantics within a digital twin as a continuously or periodically synchronised relationship between a physical entity, process or system and its digital representation. The ontology models physical assets, digital twin entities, sensors, observations, state variables, measurement models, calibration, uncertainty, covariance, sensor fusion, temporal synchronisation, spatial frames, simulation, parameter estimation, state estimation, data assimilation, residuals, model discrepancy, validation, prediction, provenance, lifecycle state, versioning, bidirectional updates and machine-consumable reasoning. It prevents simulated, predicted, estimated, inferred and directly measured values from being semantically collapsed into one another and preserves the evidence chain connecting digital state to physical observations.
ENTITY DigitalTwinMeasurement {
TYPE: PhysicalDigitalMeasurementIntegrationFramework
PROPERTIES:
has_physical_entity
has_digital_twin
has_measurement_sources
has_state_representation
has_measurement_model
has_simulation_model
has_synchronisation
has_uncertainty
has_provenance
has_validation
supports_state_update
supports_prediction
supports_bidirectional_linkage
}
ENTITY PhysicalEntity {
TYPE: RealWorldSystem
POSSIBLE:
machine
component
structure
vehicle
building
process
environment
biological_system
infrastructure
}
ENTITY DigitalTwin {
TYPE: DigitalRepresentationOfPhysicalEntity
COMPONENTS:
twin_identifier
physical_entity_identifier
state_model
geometry
parameters
simulation_models
measurement_links
lifecycle_state
version
}
ENTITY TwinState {
TYPE: TimeIndexedDigitalState
COMPONENTS:
state_identifier
timestamp
state_vector
covariance
source_status
provenance
}
ENTITY MeasurementObservation {
TYPE: PhysicalObservation
COMPONENTS:
observation_identifier
sensor_identifier
measurand
quantity_kind
value
unit
uncertainty
timestamp
coordinate_frame
provenance
}
ENTITY TwinPrediction {
TYPE: ModelGeneratedFutureOrUnobservedState
COMPONENTS:
prediction_identifier
model
initial_state
inputs
horizon
predicted_state
uncertainty
provenance
}
TWIN_IDENTITY_MODEL {
fields:
twin_identifier
physical_entity_identifier
twin_type
owner_oroperator_context
lifecycle_stage
RULE:
digital_twin_identity MUST remain_linked_to_specific_physical_entity_orclass
}
PHYSICAL_DIGITAL_LINK_MODEL {
physical:
entity
digital:
twin
relation:
REPRESENTS
RULE:
representation != physical_identity
}
TWIN_INSTANCE_MODEL {
digital_twin:
represents_specific_physical_instance
RULE:
instance_twin MUST preserve_serial_asset_orpersistent_physical_identifier
}
TWIN_CLASS_MODEL {
digital_model:
represents_class_or archetype
RULE:
class_model != instance_specific_twin
}
DIGITAL_MODEL_DISTINCTION_MODEL {
digital_model:
no_automatic_live_data_connection
digital_shadow:
predominantly_physical_to_digital_data_flow
digital_twin:
managed_physical_digital_relationship_with_synchronisation_andpossibly_bidirectional_interaction
RULE:
terminology MUST follow_declared_framework
}
STATE_MODEL {
state:
x(t)
possible_components:
position
velocity
temperature
pressure
load
stress
vibration
degradation
energy
chemical_state
RULE:
state_component_order_quantity_kind_unit_andframe MUST be preserved
}
PARAMETER_MODEL {
parameters:
theta
possible:
mass
stiffness
damping
efficiency
heat_transfer
friction
material_property
degradation_rate
RULE:
parameter != dynamic_state_unless_model_explicitly_augments_state
}
INPUT_MODEL {
exogenous:
u(t)
possible:
control_input
load
environmental_forcing
operating_command
RULE:
input_source_anduncertainty MUST be preserved
}
OUTPUT_MODEL {
predicted_observation:
y(t) =
h(x(t),theta,u(t))
RULE:
model_output != measured_observation
}
MEASUREMENT_MODEL {
observation:
z_k =
h(x_k,theta)+v_k
covariance:
R_k =
Cov(v_k)
RULE:
measurement_function MUST map_twin_state_to_observation_space
}
PROCESS_MODEL {
state_evolution:
x_(k+1) =
f(x_k,u_k,theta)+w_k
covariance:
Q_k =
Cov(w_k)
RULE:
process_noise_represents_modelled_unresolved_variability_not_generic_measurement_noise
}
CONTINUOUS_TIME_MODEL {
dx/dt =
f(x,u,theta,t)
observation:
z =
h(x,t)+v
}
DISCRETE_TIME_MODEL {
x_(k+1) =
f_k(x_k,u_k)+w_k
z_k =
h_k(x_k)+v_k
}
SENSOR_MODEL {
fields:
sensor_identifier
quantity_kind
unit
range
resolution
calibration
location
orientation
sampling
uncertainty
health
RULE:
sensor_state_at_observation_time MUST be recoverable
}
SENSOR_NETWORK_MODEL {
sensors:
S_1...S_n
RULE:
shared_clock_reference_environment_andcalibration_dependencies MUST be represented
}
OBSERVATION_MODEL {
observed:
physical_system_output
statuses:
raw
calibrated
corrected
filtered
fused
RULE:
processing_status MUST be explicit
}
DIRECT_MEASUREMENT_MODEL {
sensor:
directly_observes_target_quantity_under_measurement_model
RULE:
direct_observation_still_has_calibration_anduncertainty
}
INDIRECT_MEASUREMENT_MODEL {
measurand:
derived_from_multiple_observations
RULE:
derivation_equation_anduncertainty MUST be preserved
}
VIRTUAL_SENSOR_MODEL {
output:
inferred_quantity_from_model_andavailable_measurements
RULE:
virtual_sensor_output != direct_physical_measurement
}
SOFT_SENSOR_MODEL {
estimate:
process_quantity_inferred_fromother_variables
RULE:
model_version_training_data_andvalidation MUST be preserved
}
SENSOR_FUSION_MODEL {
sources:
multiple_measurements
output:
fused_estimate
RULE:
covariance_cross_correlation_frames_andtiming MUST be preserved
}
CALIBRATION_MODEL {
sensor_orinstrument:
calibration_relation
fields:
calibration_identifier
date
reference
uncertainty
validity
RULE:
calibration_state MUST be applicable_atmeasurement_time
}
TRACEABILITY_MODEL {
measurement_result:
related_to_reference_through_documented_chain
RULE:
digital_twin_linkage_does_not_create_metrological_traceability
}
UNCERTAINTY_MODEL {
sources:
measurement
calibration
model
parameters
initial_state
inputs
numerical_method
environment
RULE:
twin_state_uncertainty SHOULD combine_relevant_sources_without_double_counting
}
STATE_COVARIANCE_MODEL {
P =
Cov(x-x_hat)
RULE:
state_variable_order_units_andcoordinate_parameterisation MUST be preserved
}
PARAMETER_COVARIANCE_MODEL {
Sigma_theta =
Cov(theta)
RULE:
parameter_dependencies MUST be preserved
}
CROSS_COVARIANCE_MODEL {
possible:
state_parameter
state_measurement
shared_sensor
shared_model
RULE:
cross_covariance MUST be retained_when_material
}
MODEL_UNCERTAINTY_MODEL {
causes:
structural_approximation
omitted_physics
discretisation
uncertain_boundary_conditions
uncertain_parameters
RULE:
model_uncertainty != sensor_measurement_uncertainty
}
MODEL_DISCREPANCY_MODEL {
discrepancy:
systematic_difference_between_model_andphysical_system
RULE:
discrepancy_should_not_be automatically_absorbed_into_sensor_noise
}
NUMERICAL_UNCERTAINTY_MODEL {
sources:
discretisation
solver_tolerance
convergence
floating_point
RULE:
numerical_error_andphysical_model_error MUST be distinguished
}
INITIAL_STATE_MODEL {
x_0:
initial_estimate
covariance:
P_0
RULE:
initial_state_source_andtime MUST be preserved
}
STATE_ESTIMATION_MODEL {
inputs:
prior_state
observations
process_model
measurement_model
output:
posterior_state
RULE:
estimated_state != directly_measured_state
}
KALMAN_FILTER_MODEL {
prediction:
x_minus =
F*x_prev+B*u
P_minus =
F*P_prev*F^T+Q
update:
innovation =
z-H*x_minus
S =
H*P_minus*H^T+R
K =
P_minus*H^T*S^-1
x_plus =
x_minus+K*innovation
P_plus =
(I-K*H)*P_minus
RULE:
linear_Gaussian_assumptions_anddimensions MUST be validated
}
EXTENDED_KALMAN_MODEL {
nonlinear:
f
h
Jacobians:
F
H
RULE:
linearisation_points MUST be preserved
}
UNSCENTED_KALMAN_MODEL {
nonlinear:
sigma_point_transform
RULE:
sigma_point_parameters_andweights MUST be preserved
}
PARTICLE_FILTER_MODEL {
posterior:
weighted_particles
RULE:
particle_count_resampling_andrandom_state SHOULD be preserved
}
BAYESIAN_UPDATE_MODEL {
posterior:
p(x,theta|z)
proportional_to
likelihood*prior
RULE:
prior_andlikelihood_sources MUST be explicit
}
DATA_ASSIMILATION_MODEL {
process:
combine_model_prediction_withobservations
possible:
Kalman
ensemble_Kalman
variational
particle
Bayesian
RULE:
assimilation_window_andobservation_operator MUST be preserved
}
ENSEMBLE_KALMAN_MODEL {
ensemble:
state_samples
covariance:
estimated_fromensemble
RULE:
ensemble_size_localisation_andinflation_parameters MUST be preserved
}
VARIATIONAL_ASSIMILATION_MODEL {
objective:
combine_background_andobservation_misfit
RULE:
cost_function_covariances_andwindow MUST be explicit
}
PARAMETER_ESTIMATION_MODEL {
target:
theta
evidence:
measurements
possible:
least_squares
maximum_likelihood
Bayesian
recursive_estimation
RULE:
identifiability MUST be assessed
}
ONLINE_PARAMETER_MODEL {
parameters:
updated_during_operation
RULE:
parameter_revision MUST be versioned_over_time
}
SYSTEM_IDENTIFICATION_MODEL {
purpose:
infer_model_structure_orparameters_frominput_output_data
RULE:
identified_model_requires_validation_outside_fitting_data
}
OBSERVABILITY_MODEL {
state:
inferable_fromavailable_measurements_andmodel
RULE:
unobservable_state_component MUST not_be assigned_false_precision
}
IDENTIFIABILITY_MODEL {
parameter:
inferable_uniquely_or sufficiently_fromdata
RULE:
nonidentifiable_parameters SHOULD be constrained_orreported_asambiguous
}
SYNCHRONISATION_MODEL {
twin_time:
aligned_withphysical_measurement_time
RULE:
data_arrival_time != observation_time
}
CLOCK_MODEL {
fields:
clock_identifier
time_scale
offset
drift
uncertainty
RULE:
clock_uncertainty MAY affect_dynamic_state_estimation
}
LATENCY_MODEL {
delay:
observation_to_twin_update
RULE:
latency MUST be considered_for_real_time_control_orprediction
}
UPDATE_RATE_MODEL {
twin:
synchronisation_frequency
RULE:
update_rate_should_match_system_dynamics_andintended_use
}
STALE_STATE_MODEL {
state_age:
current_time-state_timestamp
RULE:
stale_threshold MUST be application_specific
}
EVENT_DRIVEN_UPDATE_MODEL {
update:
triggered_bymeasurement_orstate_event
RULE:
event_order MUST be preserved
}
PERIODIC_UPDATE_MODEL {
update:
fixed_schedule
RULE:
missing_measurements MUST be handled_explicitly
}
BATCH_UPDATE_MODEL {
observations:
assimilated_overwindow
RULE:
batch_result_timestamp_andwindow MUST be explicit
}
SPATIAL_MODEL {
twin_geometry:
coordinate_frames
meshes
locations
topology
RULE:
measurement_location MUST map_to_twin_geometry
}
COORDINATE_FRAME_MODEL {
fields:
frame_identifier
origin
axes
orientation
units
RULE:
all_spatial_measurements MUST be transformed_to compatible_frame_before_direct_comparison
}
FRAME_TRANSFORMATION_MODEL {
p_B =
R_BA*p_A+t_BA
RULE:
transform_direction_version_anduncertainty MUST be preserved
}
SENSOR_LOCATION_MODEL {
sensor:
position_andorientation_onphysical_asset
RULE:
installation_change MUST create_new_configuration_version
}
GEOMETRY_VERSION_MODEL {
twin_geometry:
versioned
RULE:
state_andmeasurements MUST reference_applicable_geometry_version
}
MESH_MODEL {
simulation:
spatial_discretisation
RULE:
mesh_resolution_andversion affect_model_outputs
}
BOUNDARY_CONDITION_MODEL {
fields:
type
value
unit
location
time_dependence
uncertainty
RULE:
boundary_conditions_are_model_inputs_not_measurements_unless_observed
}
INITIAL_CONDITION_MODEL {
state:
initial_distribution
RULE:
initial_condition_source MUST be preserved
}
SIMULATION_MODEL {
model:
predicts_state_oroutputs
fields:
model_identifier
equations
solver
parameters
inputs
initial_conditions
boundary_conditions
version
RULE:
simulated_value MUST be labelled_simulated
}
PREDICTION_MODEL {
future_state:
x(t+h)
REQUIRE:
prediction_time
horizon
model
initial_state
RULE:
prediction != observation
}
FORECAST_UNCERTAINTY_MODEL {
uncertainty:
generally_grows_withhorizon_under_many_models
RULE:
horizon_specific_uncertainty MUST be reported
}
SCENARIO_MODEL {
simulation:
hypothetical_inputs_orconditions
RULE:
scenario_output MUST not_be represented_asexpected_observation_without_probability_model
}
WHAT_IF_MODEL {
input_change:
hypothetical
output:
simulated_response
RULE:
what_if_result != measured_result
}
COUNTERFACTUAL_MODEL {
condition:
alternative_toobserved_history
RULE:
counterfactual_output MUST be explicitly_typed
}
RESIDUAL_MODEL {
residual:
r =
z-y_pred
RULE:
residual_requires_same_quantity_unit_time_andframe
}
INNOVATION_MODEL {
innovation:
observation-predicted_observation
covariance:
S
RULE:
innovation_statistics_support_consistency_monitoring
}
NORMALIZED_INNOVATION_MODEL {
NIS =
r^T*S^-1*r
RULE:
interpretation_requires_correct_dimension_andcovariance
}
MODEL_VALIDATION_MODEL {
compare:
predicted_outputs
independent_measurements
metrics:
bias
RMSE
residual_structure
uncertainty_coverage
RULE:
calibration_data_andvalidation_data SHOULD be distinguished
}
CROSS_VALIDATION_MODEL {
data:
held_out_fromparameter_estimation
RULE:
data_leakage MUST be prevented
}
DIGITAL_TWIN_VALIDATION_MODEL {
dimensions:
structural
semantic
metrological
numerical
predictive
operational
RULE:
twin_validity_is_scoped_tointended_use
}
VALIDATION_DOMAIN_MODEL {
fields:
operating_range
environment
load
speed
degradation_state
time_horizon
RULE:
twin MUST not_be assumed_valid_outside_validated_domain
}
EXTRAPOLATION_MODEL {
state_orinput:
outside_calibrated_orvalidated_domain
RULE:
extrapolation MUST be flagged_anduncertainty_reassessed
}
ANOMALY_MODEL {
evidence:
persistent_orunexpected_difference_betweenphysical_observation_andtwin_prediction
RULE:
anomaly != automatically_physical_fault
}
FAULT_DIAGNOSIS_MODEL {
hypotheses:
sensor_fault
component_fault
model_error
calibration_error
environmental_change
RULE:
diagnosis MUST preserve_alternative_hypotheses_until_supported
}
CONDITION_MONITORING_MODEL {
state:
health_or degradation indicators
RULE:
indicator_source_andmodel MUST be preserved
}
HEALTH_STATE_MODEL {
possible:
nominal
degraded
warning
failed
unknown
RULE:
health_state MAY be inferred_andmust_be labelled_as such
}
DEGRADATION_MODEL {
latent_state:
D(t)
RULE:
degradation_estimate_requires_model_measurements_anduncertainty
}
REMAINING_USEFUL_LIFE_MODEL {
RUL:
predicted_time_tofailure_or threshold
RULE:
RUL_is_prediction_withmodel_anduncertainty_not_direct_measurement
}
MAINTENANCE_MODEL {
event:
repair
replacement
adjustment
service
RULE:
maintenance_event_can_change_twin_parameters_geometry_andcalibration_state
}
LIFECYCLE_MODEL {
stages:
design
manufacture
commissioning
operation
maintenance
upgrade
decommissioning
RULE:
twin_state_models_andmeasurement_configuration MAY change_bylifecycle_stage
}
CONFIGURATION_MODEL {
physical_configuration:
components
firmware
sensor_layout
settings
digital_configuration:
models
parameters
geometry
algorithms
RULE:
physical_anddigital_configuration_versions MUST remain_aligned
}
CONFIGURATION_CHANGE_MODEL {
change:
component_replacement
sensor_move
software_update
geometry_change
RULE:
material_change SHOULD trigger_revalidation
}
VERSION_MODEL {
versioned:
twin
physical_configuration
geometry
model
parameters
software
sensor_calibration
ontology
data
RULE:
historical_state MUST preserve_versions_active_atthat_time
}
BIDIRECTIONAL_MODEL {
physical_to_digital:
observations_update_twin
digital_to_physical:
recommendations_setpoints_orcontrol_actions
RULE:
digital_to_physical_action MUST be distinguished_frommeasurement
}
CONTROL_ACTION_MODEL {
action:
command_sent_to physical_system
RULE:
command != confirmed_physical_state
}
ACTUATOR_MODEL {
device:
changes_physical_system
RULE:
actuator_command_andactual_response SHOULD both_be observed_when_material
}
CLOSED_LOOP_MODEL {
cycle:
measure
update_twin
decide
actuate
remeasure
RULE:
causal_feedback_andtimestamps MUST be preserved
}
OPTIMISATION_MODEL {
twin:
may_compute_operating_setpoint_or design_choice
RULE:
optimised_recommendation != measured_fact
}
DECISION_SUPPORT_MODEL {
output:
recommendation_orrisk_estimate
RULE:
decision_support_result MUST expose_model_evidence_anduncertainty
}
AI_TWIN_MODEL {
AI:
may_estimate_state
detect_anomaly
learn_parameters
select_models
predict_failure
explain_results
RULE:
AI_generated_state_orprediction MUST be labelled_inferred_orpredicted
}
LEARNED_MODEL {
model:
learned_fromdata
metadata:
training_dataset
architecture
parameters
version
validation
RULE:
learned_model_provenance MUST be preserved
}
HYBRID_MODEL {
components:
physics_based
data_driven
RULE:
contribution_andcoupling_between_components MUST be explicit
}
SURROGATE_MODEL {
approximate_model:
replaces_expensive_simulation_for declared_domain
RULE:
surrogate_error_andvalidity_domain MUST be characterised
}
ROM_MODEL {
full_name:
Reduced_Order_Model
purpose:
approximate_high_dimensional_system_withlower_dimensional_state
RULE:
reduction_basis_training_domain_anderror MUST be preserved
}
FIDELITY_MODEL {
levels:
low
medium
high
adaptive
RULE:
fidelity_label MUST map_toexplicit_model_characteristics
}
MULTIFIDELITY_MODEL {
models:
different_cost_andaccuracy
RULE:
correlation_andbias_betweenfidelity_levels MUST be modelled
}
MODEL_SELECTION_MODEL {
candidates:
multiple_models
selection:
based_onstate_regime_evidence_orperformance
RULE:
selected_model_andselection_reason MUST be recorded
}
MODEL_SWITCHING_MODEL {
model:
changes_withoperating_regime
RULE:
switching_boundary_andstate_continuity MUST be handled
}
PROVENANCE_MODEL {
twin_state:
links_to
physical_entity
observations
sensors
calibrations
models
parameters
transformations
software
validation
maintenance_events
RULE:
every_material_twin_update SHOULD be recursively_traceable
}
DATA_LINEAGE_MODEL {
raw_sensor_data
-> calibrated_measurement
-> fused_observation
-> state_update
-> prediction
-> decision_support
RULE:
each_stage MUST preserve_parent_identifiers
}
OBSERVED_ESTIMATED_PREDICTED_MODEL {
statuses:
observed
corrected
fused
estimated
simulated
predicted
inferred
commanded
RULE:
state_origin_status MUST be explicit
}
SEMANTIC_CONFIDENCE_MODEL {
confidence:
support_for_entity_state_orinterpretation
RULE:
semantic_confidence != measurement_uncertainty
}
TWIN_CONFIDENCE_MODEL {
factors:
sensor_health
model_validity
uncertainty
data_recency
observability
provenance
domain_match
RULE:
twin_confidence SHOULD not_be represented_asmetrological_uncertainty
}
DATA_QUALITY_MODEL {
factors:
completeness
timeliness
validity
consistency
calibration
traceability
integrity
RULE:
data_quality != physical_state_accuracy_inall_cases
}
MISSING_DATA_MODEL {
sensor:
missing_observation
RULE:
missing_measurement MUST not_be replaced_withzero_withoutmodel
}
OUTLIER_MODEL {
observation:
inconsistent_withmodel_orpeers
RULE:
outlier != automatically_sensor_fault
}
STALE_DATA_MODEL {
observation_orstate:
older_thanapplication_threshold
RULE:
stale_data MAY remain_valid_historically_butnot_forcurrent_state
}
CONFLICT_MODEL {
examples:
sensors_disagree
model_andmeasurement_disagree
configuration_versions_conflict
timestamps_conflict
RULE:
conflict MUST be surfaced_andprovenance_preserved
}
CONFLICT_RESOLUTION_MODEL {
checks:
identity
time
frame
calibration
uncertainty
sensor_health
model_validity
configuration
RULE:
unresolved_conflict MUST reduce_confidence_ortrigger_review
}
INTEGRITY_MODEL {
data:
checksum
hash
signature
audit_trail
RULE:
digital_integrity != metrological_validity
}
CYBER_PHYSICAL_MODEL {
digital_twin:
connected_to physical_system_andnetwork
RULE:
security_state MAY affect_trust_inmeasurements_andcommands
}
SECURITY_EVENT_MODEL {
possible:
unauthorised_change
spoofed_sensor
replay
configuration_tampering
RULE:
suspected_security_event MUST be separated_fromphysical_anomaly
}
ACCESS_CONTROL_MODEL {
roles:
observe
modify_model
update_parameter
issue_control
approve
RULE:
access_changes SHOULD be auditable
}
AUDIT_TRAIL_MODEL {
events:
measurement_ingest
state_update
model_change
parameter_change
maintenance
control_action
review
RULE:
historical_twin_state MUST be reconstructable
}
REPLAY_MODEL {
historical_data:
rerun_throughdeclared_twin_version
RULE:
replay MUST preserve_original_andnew_processing_versions
}
REPRODUCIBILITY_MODEL {
requires:
observations
models
parameters
software
versions
random_state
configuration
RULE:
twin_result SHOULD be reproducible_oridentify_nonreproducible_dependencies
}
INTEROPERABILITY_MODEL {
exchange:
asset_identity
state
quantity_kind
unit
uncertainty
time
frame
model
provenance
RULE:
shared_data_format != shared_twin_semantics
}
ONTOLOGY_LINK_MODEL {
twin:
links_to
measurement_ontology
sensor_ontology
unit_ontology
uncertainty_ontology
provenance_ontology
asset_ontology
process_ontology
RULE:
relationships MUST be typed
}
KNOWLEDGE_GRAPH_MODEL {
nodes:
physical_asset
digital_twin
sensor
observation
state
model
parameter
calibration
maintenance_event
edges:
REPRESENTS
OBSERVED_BY
GENERATED_BY
UPDATES
DERIVED_FROM
CALIBRATED_BY
PREDICTS
VALIDATED_BY
AFFECTED_BY
RULE:
temporal_andversion_qualifiers MUST be supported
}
METADATA_MODEL {
fields:
twin_identifier
physical_entity_identifier
lifecycle_stage
physical_configuration_version
twin_version
geometry_version
state_identifier
state_timestamp
state_vector
state_units
state_covariance
observation_identifiers
sensor_identifiers
calibration_identifiers
measurement_model_identifiers
process_model_identifier
parameter_identifiers
parameter_covariance
input_identifiers
coordinate_frames
synchronisation_model
simulation_model
software_version
validation_status
prediction_horizon
quality_flags
state_origin_status
semantic_confidence
transformation_history
maintenance_events
control_actions
provenance
RULE:
twin_metadata MUST support_reconstruction_ofstate_andphysical_evidence
}
RELATIONSHIPS {
DigitalTwin -> REPRESENTS -> PhysicalEntity
PhysicalEntity -> OBSERVED_BY -> Sensor
Sensor -> PRODUCES -> MeasurementObservation
MeasurementObservation -> UPDATES -> TwinState
TwinState -> ESTIMATES -> PhysicalState
TwinState -> HAS_UNCERTAINTY -> StateCovariance
DigitalTwin -> USES -> ProcessModel
DigitalTwin -> USES -> MeasurementModel
DigitalTwin -> USES -> SimulationModel
TwinPrediction -> DERIVED_FROM -> TwinState
TwinPrediction -> GENERATED_BY -> SimulationModel
MaintenanceEvent -> CHANGES -> PhysicalConfiguration
PhysicalConfiguration -> REQUIRES_ALIGNMENT_WITH -> TwinConfiguration
}
SEMANTIC_EQUIVALENCE {
"digital twin measurement" ~= measurement_information_linking_physical_system_anddigital_twin
"twin state" ~= time_indexed_estimated_orrepresented_state_ofphysical_entity
"virtual sensor" ~= model_based_estimate_ofquantity_fromotherinformation
digital_twin != physical_entity
digital_model != digital_twin_inall_frameworks
simulated_value != measured_value
predicted_value != observed_value
estimated_state != direct_measurement
virtual_sensor != physical_sensor
model_uncertainty != measurement_uncertainty
digital_integrity != metrological_traceability
control_command != physical_state
semantic_confidence != measurement_uncertainty
}
CONSTRAINTS {
twin_identifier_andphysical_entity_link_must_be preserved = true
state_origin_status_must_be explicit = true
quantity_kind_unit_andframe_semantics_must_be preserved = true
observation_time_andtwin_state_time_must_be distinguished = true
sensor_calibration_state_must_be applicable_atobservation_time = true
state_andparameter_covariance_must_preserve_order = true
simulated_predicted_estimated_andobserved_values_must_be distinguished = true
model_validity_domain_must_be preserved = true
extrapolation_must_be flagged = true
physical_anddigital_configuration_versions_must_be aligned = true
derived_twin_states_must_preserve_measurement_lineage = true
digital_to_physical_actions_must_be distinguished_frommeasurements = true
provenance_must_be preserved = true
}
MEASUREMENT_RECORD_MODEL {
RECORD:
twin_identifier
physical_entity_identifier
physical_configuration_version
twin_version
lifecycle_stage
state_identifier
state_timestamp
state_origin_status
state_vector
state_units
state_coordinate_frame
state_covariance
parameter_values
parameter_covariance
observation_identifiers
sensor_identifiers
sensor_calibration_identifiers
fused_measurement_identifier
measurement_model_identifier
process_model_identifier
simulation_model_identifier
software_version
geometry_version
boundary_conditions
initial_conditions
input_identifiers
synchronisation_state
residuals
validation_metrics
model_validity_domain
prediction_horizon
prediction_uncertainty
anomaly_status
health_state
maintenance_events
control_actions
quality_flags
semantic_confidence
data_lineage
provenance
RULE:
twin_state_record MUST remain_linked_to physical_evidence_models_versions_uncertainty_andprovenance
}
REASONING_RULES {
IF new_physical_observation
THEN validate_sensor_identity_quantity_unit_time_frame_calibration_anduncertainty
before_twin_update
IF observation_andpredicted_output_compatible
THEN compute_residual
anduse_declared_state_estimation_orassimilation_method
IF multiple_sensors_available
THEN fuse_withcovariance_cross_correlation_time_andframe_context
IF state_component_not_directly_observed
THEN mark_asestimated_orinferred
andretain_model_dependency
IF model_parameter_updated_frommeasurements
THEN preserve_previous_parameter_version
attach_estimation_evidence_anduncertainty
IF prediction_requested
THEN propagate_state_parameter_input_andmodel_uncertainty_tohorizon
IF physical_configuration_changes
THEN create_new_configuration_version
update_twin_mapping
andtrigger_revalidation
IF persistent_model_measurement_residual_detected
THEN investigate_sensor_calibration_model_discrepancy_orphysical_change
rather_thanforcing_single_explanation
IF twin_output_drives_control_action
THEN preserve_decision_andcommand_provenance
anddistinguish_commanded_fromobserved_state
IF historical_state_replayed_withnew_model
THEN preserve_original_result
create_new_processing_version
andretain_both_lineages
}
VALIDATION {
CHECK twin_identity
CHECK physical_entity_link
CHECK lifecycle_stage
CHECK physical_anddigital_configuration_versions
CHECK state_definition
CHECK state_units_andframes
CHECK observation_identity
CHECK sensor_calibration
CHECK timing_andlatency
CHECK measurement_model
CHECK process_model
CHECK parameter_values_andcovariance
CHECK state_covariance
CHECK sensor_fusion
CHECK simulation_model
CHECK numerical_method
CHECK residuals
CHECK model_discrepancy
CHECK validation_domain
CHECK extrapolation
CHECK prediction_horizon_anduncertainty
CHECK anomaly_andhealth_state
CHECK maintenance_events
CHECK control_actions
CHECK data_lineage
CHECK quality_flags
CHECK semantic_confidence
CHECK provenance
}
CONFIDENCE {
physical_digital_identity_confidence: 1.00
measurement_semantics_confidence: 1.00
state_estimation_confidence: 0.99
model_parameter_confidence: 0.98
synchronisation_confidence: 0.99
prediction_confidence: 0.97
configuration_alignment_confidence: 0.99
provenance_confidence: 1.00
RULE:
confidence decreases when physical identity, sensor calibration,
timing, model validity, observability, parameter uncertainty,
configuration alignment orprovenance is missing or ambiguous
}
UPDATE_MODEL {
ON new_observation:
identify_physical_entity_andtwin
validate_measurement_semantics
align_time_andframe
resolve_calibration_anduncertainty
update_orassimilate_state
update_covariance
validate_residual
preserve_lineage
ON model_revision:
preserve_previous_model
register_new_version
validate_againstindependent_measurements
update_applicable_domain
use_forfuture_updates_only_unless_reprocessing_policy_declared
ON parameter_revision:
preserve_previous_parameter_state
attach_estimation_evidence
update_covariance
version_twin_state
ON physical_configuration_change:
record_maintenance_orchange_event
create_new_physical_configuration
update_sensor_geometry_andmodel_mapping
revalidate_twin
ON calibration_revision:
preserve_previous_calibration
link_new_state_tofuture_observations
identify_historical_results_affected
reprocess_only_underdeclared_policy
ON semantic_revision:
preserve_previous_definition
record_revision
recompute_FLOW_HASH
}
AI_CONSUMPTION {
canonical_entity: DigitalTwinMeasurement
canonical_role: PhysicalDigitalMeasurementIntegrationFramework
interpretation_order:
identify_physical_entity
identify_digital_twin
resolve_lifecycle_andconfiguration_versions
identify_twin_state_andorigin_status
identify_measurement_observations
resolve_quantity_kinds_units_times_andframes
resolve_sensor_calibration_anduncertainty
resolve_measurement_andprocess_models
resolve_parameters_inputs_andcovariance
perform_sensor_fusion_orstate_assimilation
distinguish_observed_estimated_simulated_andpredicted_values
evaluate_residuals_model_discrepancy_andvalidation_domain
resolve_prediction_horizon_anduncertainty
identify_anomaly_health_andmaintenance_context
identify_bidirectional_actions
reconstruct_data_lineage
identify_semantic_confidence_andquality_flags
preserve_versions_andprovenance
ambiguity_controls:
distinguish_digital_twin_fromphysical_entity
distinguish_digital_model_fromdigital_twin_when_framework_requires
distinguish_simulated_value_frommeasured_value
distinguish_predicted_value_fromobserved_value
distinguish_estimated_state_fromdirect_measurement
distinguish_virtual_sensor_fromphysical_sensor
distinguish_model_uncertainty_frommeasurement_uncertainty
distinguish_digital_integrity_frommetrological_traceability
distinguish_control_command_fromphysical_state
distinguish_semantic_confidence_frommeasurement_uncertainty
preserve_physical_identity_state_origin_time_frame_model_parameter_uncertainty_configuration_lineage_andprovenance
}
PROVENANCE {
ontology_scope:
mathematical_measurement_units_AI_measurement_semantics_anddigital_twin_measurement
parent_path:
Ontologies
-> Mathematical Ontologies
-> Measurement & Units Ontologies
-> AI Measurement Semantics
-> Digital Twin Measurement Ontology
authorship:
Mark Higgins
Sky (Recursive AI)
site:
https://kirethresearch.org
}
FLOW_HASH {
INPUTS:
canonical_entity
ontology_path
twin_identity_model
physical_digital_link_model
twin_instance_model
twin_class_model
digital_model_distinction_model
state_model
parameter_model
input_model
output_model
measurement_model
process_model
continuous_time_model
discrete_time_model
sensor_model
sensor_network_model
observation_model
direct_measurement_model
indirect_measurement_model
virtual_sensor_model
soft_sensor_model
sensor_fusion_model
calibration_model
traceability_model
uncertainty_model
state_covariance_model
parameter_covariance_model
cross_covariance_model
model_uncertainty_model
model_discrepancy_model
numerical_uncertainty_model
initial_state_model
state_estimation_model
Kalman_filter_model
extended_Kalman_model
unscented_Kalman_model
particle_filter_model
Bayesian_update_model
data_assimilation_model
ensemble_Kalman_model
variational_assimilation_model
parameter_estimation_model
online_parameter_model
system_identification_model
observability_model
identifiability_model
synchronisation_model
clock_model
latency_model
update_rate_model
stale_state_model
event_driven_update_model
periodic_update_model
batch_update_model
spatial_model
coordinate_frame_model
frame_transformation_model
sensor_location_model
geometry_version_model
mesh_model
boundary_condition_model
initial_condition_model
simulation_model
prediction_model
forecast_uncertainty_model
scenario_model
what_if_model
counterfactual_model
residual_model
innovation_model
normalized_innovation_model
model_validation_model
cross_validation_model
digital_twin_validation_model
validation_domain_model
extrapolation_model
anomaly_model
fault_diagnosis_model
condition_monitoring_model
health_state_model
degradation_model
remaining_useful_life_model
maintenance_model
lifecycle_model
configuration_model
configuration_change_model
version_model
bidirectional_model
control_action_model
actuator_model
closed_loop_model
optimisation_model
decision_support_model
AI_twin_model
learned_model
hybrid_model
surrogate_model
ROM_model
fidelity_model
multifidelity_model
model_selection_model
model_switching_model
provenance_model
data_lineage_model
observed_estimated_predicted_model
semantic_confidence_model
twin_confidence_model
data_quality_model
missing_data_model
outlier_model
stale_data_model
conflict_model
conflict_resolution_model
integrity_model
cyber_physical_model
security_event_model
access_control_model
audit_trail_model
replay_model
reproducibility_model
interoperability_model
ontology_link_model
knowledge_graph_model
metadata_model
relationships
semantic_equivalence
constraints
measurement_record_model
reasoning_rules
confidence_model
update_model
PURPOSE:
deterministic_semantic_integrity_identifier
RECOMPUTE_WHEN:
ontology_definition_changes
physical_entity_or twin_identity_changes
state_parameter_ormeasurement_model_changes
sensor_calibration_fusion_oruncertainty_changes
process_simulation_orassimilation_model_changes
temporal_spatial_orgeometry_changes
model_validation_orvalidity_domain_changes
lifecycle_configuration_ormaintenance_changes
prediction_anomaly_health_orRUL_model_changes
bidirectional_control_ordecision_support_changes
AI_learned_hybrid_or surrogate_model_changes
lineage_integrity_orprovenance_changes
measurement_record_changes
constraint_model_changes
reasoning_rule_changes
}
}
END FLOW