A Semantic Approach to Spacecraft Verification Planning Using Bayesian Networks

The design and execution of an effective spacecraft verification strategy is a critical and complex undertaking. It necessitates the generation and management of a large amount of highly connected data. As a result, verification activities consume a significant part, if not the biggest part, of the development costs of large-scale engineered systems. Current document-based approaches to verification planning and assessment are inefficient, prone to inconsistencies, and unable to quantitatively inform about the confidence level on the verification status of the system of interest. A Digital Engineering (DE) approach to verification planning and assessment can overcome these weaknesses. To achieve this, data integration and management technologies are crucial. Semantic Web Technologies (SWTs) provide a means to structure knowledge, validate knowledge, and infer new knowledge using ontologies, reasoners, and query languages. In this paper, we present a Bayesian approach to planning verification strategies supported by the Bayesian Verification Ontology Stack (BVOS). The BVOS is a modular ontology stack that supports a semantic approach to DE. The ontologies it comprises are constructed using the Ontological Modeling Language (OML). The BVOS leverages existing ontologies such as the Basic Formal Ontology (BFO) and the Common Core Ontologies (CCO) to support the required domain-level ontologies such as the System Architecture Ontology and the Bayesian Network Ontology. To evaluate this approach, it has been applied to the Attitude Determination and Control Subsystem (ADCS) of a notional spacecraft and its verification strategy. The ADCS requirements are captured in Jama Connect. The physical architecture of the ADCS and the corresponding verification strategy are modeled using the Systems Modeling Language (SysML) v2. A representative OML knowledge graph that captures the entire dataset is produced and validated using the BVOS, and is used to generate a Bayesian representation of the verification strategy.