Industrial robots deliver high precision in structured environments but remain vulnerable to unexpected changes in sensing, dynamics, or task constraints. Conventional industrial control architectures rely on fixed policies and static perception–action mappings, limiting their ability to adapt in evolving production settings. In contrast, motivational learning mechanisms have shown promise for open-ended adaptation in laboratory robotics. This paper bridges these domains by introducing a motivational engine designed to enhance industrial robot resilience under operational disruptions. The proposed framework integrates multiple intrinsic drives—novelty, curiosity, competence, frustration, and effectance—into a unified action-selection mechanism that dynamically balances exploration and skill refinement. We evaluate the approach in a manipulation task involving observation-frame distortions and environmental changes that invalidate previously successful behaviors, demonstrating autonomous recovery from sensorimotor disruptions and the discovery of alternative kinematic strategies.