Learning in Electrical and Mechatronic Engineering as a Communication System: A Systems-Theoretic Reframing

Persistent learning difficulties in electrical and mechatronic engineering suggest a structural mismatch between the organization of disciplinary knowledge and prevailing instructional models. This paper proposes a systems-theoretic reframing in which learning is modeled explicitly as a communication and control system. Instructional content is treated as a transmitted signal, the learner as a noisy, bandwidth-limited adaptive subsystem, and understanding as successful decoding that enables prediction and control of physical processes. Drawing on Shannon information theory, Wiener's cybernetics, Kalman estimation, and predictive coding, learning is formalized as the transmission, compression, and stabilization of temporally structured information under constraints of noise, delay, and cost. Within this framework, commonly reported learning failures emerge as consequences of channel mismatch, suboptimal encoding, and poorly designed feedback. The formulation yields measurable system-level quantities, including mutual information, prediction error, entropy reduction, and stability margins, and supports experimentally testable human-in-the-loop designs grounded in systems, communication and control principles.