Simulation- and AI-supported technical troubleshooting training for biomedical device technology students in intensive care equipment education

BACKGROUND: Biomedical device technology students need opportunities to understand how intensive care equipment behaves under normal and fault conditions. However, many educational approaches in this area still rely on passive materials and offer limited opportunities for students to observe dynamic device behavior, waveform changes, alarm patterns, and system-level fault mechanisms in a safe, repeatable way. METHODS: This study presents ICU Simulator, a browser-based, multi-device simulation platform designed to support mechanism-based technical troubleshooting and structured interpretation of intensive care device behavior. The platform integrates deterministic simulation models for mechanical ventilators, patient monitors, infusion pumps, and CRRT systems, enabling real-time visualization of waveform dynamics and device behavior. In addition, a hybrid AI-supported interpretation module combines rule-based interpretation with optional large language model (LLM)-supported reformulation to provide structured feedback, possible device-related fault mechanisms, suggested verification steps, and teaching notes. The platform was examined through an exploratory pilot implementation with second-year biomedical device technology students. RESULTS: The pilot implementation generated preliminary descriptive findings on how students interacted with the simulator and approached device-related fault scenarios. Instructor observations, student scenario responses, and short post-session feedback suggested that some students began to move from output-oriented descriptions toward more mechanism-based explanations of device behavior. The AI-supported module was used mainly as a reflective support tool, helping learners compare their technical fault hypotheses with structured feedback rather than providing direct answers. CONCLUSIONS: ICU Simulator may serve as a scalable and accessible preparatory learning environment for intensive care equipment education. By integrating real-time simulation with AI-supported cognitive scaffolding, the platform may support technical training by helping learners develop structured technical reasoning, waveform interpretation, and first-line troubleshooting skills relevant to intensive care equipment education. Because the study was exploratory and small-scale, the findings should be interpreted as preliminary observations rather than definitive evidence of educational effectiveness.