RoSO meets SMGI: Formal guarantee for safe robot service reconfiguration

The paper addresses a critical gap in service robotics: while ontology conformance ensures that a robotic service description is well-formed, it fails to guarantee that reconfigurations (rebinding, recomposition, repair, redeployment) produce an admissible realization of the same protected service. Aomar Osmani argues that the Structural Model of General Intelligence (SMGI) is uniquely suited to fill this gap. SMGI provides not only a structural interface θ but also an induced behavioral semantics Tθ and a governance discipline for norm-respecting change. By embedding RoSO into SMGI as a typed semantic layer, service descriptions become dynamically governable—meaning changes can be evaluated for permissibility against the original service intent, not just syntactic correctness.

The core contribution is a RoSO-to-SMGI adequacy theorem that formalizes when a reconfigured service remains identity-preserving. The paper also defines compositional conditions under which locally acceptable updates (e.g., swapping a component) remain globally admissible across the entire service architecture. This moves beyond static conformance to a runtime governance framework. The result is not a replacement of RoSO but a formal account of admissible change. For practitioners, this means robotic systems can be updated with confidence that the updated service still meets its original semantic constraints—critical for safety-critical and long-lived autonomous systems.