Making sense of incomplete and conflicting narrative knowledge in the presence of abnormalities, unobservable processes, and other real world considerations is a challenge andcrucial requirement for cognitive robotics systems. An addedchallenge, even when suitably specialised action languagesand reasoning systems exist, is practical integration and application within large-scale robot control frameworks.In the backdrop of an autonomous wheelchair robot control task, we report on application-driven work to realisepostdiction triggered abnormality detection and re-planningfor real-time robot control: (a). Narrative-based knowledgeabout the environment is obtained via a larger smart environment framework; and (b) Abnormalities are postdictedfrom stable-models of an answer-set program correspondingto the robot's epistemic model. The overall reasoning is performed in the context of an approximate epistemic action theory based planner implemented via a translation to answer-setprogramming.
