Application-Oriented Learning in Engineering

A Retrospective Study on Teaching Cyber-Physical Intralogistics Systems

Abstract

Engineering education faces declining enrolments, international competition, and the challenge of AI solving traditional exam tasks. Intralogistic systems link mechatronics and algorithms, offering tangible challenges for application-oriented learning. This paper introduces a classification framework based on Input, Application, and Examination, and applies it to evaluate three representative courses at the Karlsruhe Institute of Technology. The analysis shows that freedom in learning, such as self-study phases, requires corresponding structure in assessment to ensure knowledge acquisition. Team-based tasks benefit from clearly defined roles and organizational scaffolding to prevent conflicts and unequal participation. The workload distribution strongly depends on course duration: very compact formats leave little room for catch-up, while extended formats require intermediate milestones to maintain motivation. Grading remains a particular challenge, as examinations must balance fairness with recognition of teamwork and practical achievements. Finally, the growing role of artificial intelligence introduces both risks and opportunities: while AI can reduce the need for routine coding, it creates new demands for creative, critical, and system-level tasks. These findings provide practical guidance for designing and evaluating interactive courses also in other areas then intralogistics.