Colloquia

Summary

Cyber-Physical Systems (CPS), such as internal logistics and warehousing systems, are increasingly being developed through platform-based design. Engineers analyze stakeholder requirements, determine capabilities, and select the subsystems that provide those capabilities based on available descriptions of predefined existing modules. In this process, three interrelated difficulties arise: (1) requirements are not represented in a format suited for configure-to-order design activities; (2) capabilities and functions of configuration items lack a formalized representation that enables consistent reuse; and (3) the configuration phase is human labor-intensive, and subject to biased interpretations. These difficulties lead to long development lead times and challenges in establishing reliable connections between requirements and system capabilities. Consequently, matching stakeholder requirements with (sub-)system capabilities, and determining candidate combinations of configuration items that best satisfy those requirements, remain bottlenecks affecting design quality. To address these challenges, this study proposes a Configuration and Requirements Processing Semantic Alignment System (CORPS) for semantic matching between requirements and capabilities. First, CORPS reformats stakeholder requirement descriptions into a machine-interpretable representation that is aligned with the capability descriptions of available configuration items. Second, CORPS provides a structured and standardized description of configuration items, enabling them to be consistently reused. Third, CORPS performs semantic matching between processed requirements and capabilities of configuration items, and identifies the candidate configuration items that best satisfy requirements. Verification and validation results from a warehouse case study show that CORPS supports a more traceable and reviewable requirement-to-capability alignment process, shifts engineering work from exhaustive processing to selective verification, and shortens the end-to-end analysis time for 80 requirements.