Reasoner

A formal, explicit specification of a shared conceptualization that defines the types, properties, and interrelationships of entities within a manufacturing domain. The reasoner uses the ontology as its rulebook, applying logical axioms to infer new facts. For example, an ontology might define that a CentrifugalPump is a subclass of RotatingEquipment, and that RotatingEquipment requires VibrationMonitoring. The reasoner then automatically classifies a new centrifugal pump as requiring vibration monitoring without explicit assertion.

A semantic web language designed to represent rich and complex knowledge about things and their relations. OWL provides the formal logic constructs—such as transitive properties, inverse relationships, and property restrictions—that a reasoner requires to perform deductive inference. Unlike basic RDF, OWL enables a reasoner to detect inconsistencies in a manufacturing knowledge graph, such as a component being classified as both a Sensor and an Actuator when the ontology defines them as disjoint classes.

A W3C standard for validating RDF graphs against a set of conditions called shapes. While a reasoner derives new knowledge, SHACL ensures the integrity of the existing data before inference occurs. For example, a SHACL shape can enforce that every MaintenanceTask node must have exactly one performedOn relationship to an Asset node. This pre-validation prevents the reasoner from propagating errors across the graph, ensuring that inferred facts are built on a trustworthy foundation.

A directed acyclic graph that encodes cause-and-effect relationships between manufacturing variables. Unlike a standard knowledge graph that asserts static facts, a causal graph enables a reasoner to perform counterfactual reasoning—simulating what would happen if a process parameter were changed. For instance, a causal reasoner can infer that increasing coolant flow rate reduces tool wear probability, enabling prescriptive recommendations rather than just diagnostic classifications.

The fundamental data structure of a knowledge graph consisting of a subject-predicate-object statement. A reasoner operates by chaining these triples together using logical rules. For example, from the triples Pump-23 hasPart Bearing-A and Bearing-A exhibitsPattern Spalling, a reasoner applying a transitive hasFailureMode property can infer the new triple Pump-23 hasFailureMode Spalling. This atomic structure is the raw material for all inference.

A knowledge graph that explicitly models the time dimension of facts, allowing a reasoner to analyze sequences and state changes. In manufacturing, a temporal reasoner can infer that a PressureSpike event preceded a SealFailure event by 15 minutes, establishing a causal chain across time. This capability is critical for root cause analysis, where the order of events is as important as the events themselves.