Bridging the STEM Divide: An AI-Assisted Approach to Developing Phyphox-Based Experiments with Low-Cost Sensors

High Resolution Image Download MS PowerPoint Slide Digital experiments play an important role in STEM education but remain underused in resource-constrained rural schools due to limited equipment and insufficient technical training. Phyphox provides a low-cost platform for digital experimentation; however, the integration of external sensors typically requires programming skills that many teachers lack. This study employed a five-stage framework, including investigation, specification, design, construction, and testing, to develop a “Generative Artificial Intelligence (GenAI) + Phyphox” model for experiment design. A CO 2 monitoring station was developed as a case study, with GenAI supporting sensor control, code generation, and data visualization. In classroom implementation, the station produced measurement trends consistent with a Pasco CO 2 sensor, and four teachers without prior technical experience successfully replicated the station. These findings indicate that this approach enables teachers to develop low-cost digital experiments suitable for qualitative and semiquantitative secondary chemistry investigations, thereby lowering technical barriers and expanding access to STEM instruction in under-resourced schools.