ERROR-CONTINGENT AI GUIDANCE AND INDEPENDENT PROBLEM SOLVING IN METROLOGY EDUCATION

An instructional procedure is proposed for moving textile engineering students from error-specific artificial intelligence (AI) guidance to independent metrological reasoning. An initial unaided attempt is followed by a targeted question, source verification and a justified revision. Support is reduced according to performance on new tasks, with provision for restoring it when a difficulty persists. A waste-composition example distinguishes a reported percentage from an unsupported claim about production losses. Correct and flawed recommendations are included in the same exercise set to assess evidence-based judgement. The contribution is a teaching procedure and an assessment proposal; no classroom trial or learning gains are reported.