Research & working papers
Engineering answers “how”; research asks “what actually works?” Everything below is verifiable: DOIs on Zenodo, live products as testbeds, or open-source code.
Flagship study
Predictive Maintenance with LSTM on the Edge
First-author study: LSTM vs GRU on IoT sensor streams for failure prediction — noise-robust training, dropout regularization and TensorFlow Lite quantization for sub-50 ms inference on edge hardware.
Active research directions
Working papers — course-based survey studies
Unpublished manuscripts from undergraduate coursework — the research-methods groundwork behind the directions above. Presented honestly as working papers, each with a DOI on Zenodo.
First author · survey of 59 pharmacy students — descriptive statistics & cross-tabulation (Excel, SPSS)
Survey of 50 students & 6 teachers — the classroom-scale groundwork behind the live StudyForce experiments (R-04)
Survey of 57 students · 22-item questionnaire on mental-health and academic impact
Looking for the smaller builds that used to live here? They're on the projects page now →