LuxRest

Light affects more than vision: its timing, intensity, and spectrum help regulate circadian rhythms. LuxRest turns near-eye light measurements into a participant-facing feedback loop. The project joins dissertation research, embedded systems, cloud services, dashboards, surveys, and messaging in one field-deployable workflow.

Wearable-light studies often end with passive measurement and retrospective analysis. LuxRest asks whether corneal-plane measurements can be interpreted and returned quickly enough to support reflection and behavior during everyday life. The work focuses first on feasibility: can the system collect, move, interpret, and communicate data reliably outside the laboratory?

The dissertation uses a within-participant crossover structure linking repeated measurement periods, participant feedback, and self-reported outcomes. The protocol brings light exposure, sleep, mood, productivity, and user experience into a shared timeline while preserving the distinction between operational feasibility and intervention efficacy.

The sensing layer centers on the LiDo light dosimeter positioned near the eye, where exposure is more meaningful than a desk-level reading. Companion electronics built around a Pico W and Raspberry Pi support local communication and transfer. My role included system requirements, integration, field preparation, and collaboration on the companion electronics and MQTT firmware.

LuxRest is an end-to-end system rather than a single device. Measurements move from the wearable through edge hardware and MQTT into cellular or cloud services, automation workflows, dashboards, surveys, and participant messaging. The architecture was designed so each layer could be inspected independently during field troubleshooting.

The pipeline aligns sensor records with study periods and participant inputs, then prepares recent exposure for visualization and feedback. Notecard and Notehub, n8n, Ubidots, Qualtrics, and Telegram support transport, orchestration, display, survey capture, and communication.

The interface reduces a complex circadian-light signal to a small number of readable cues: recent trends, time-aware context, and a traffic-light-style gauge. Feedback is written as guidance rather than diagnosis. The design goal is to help participants understand what the system observed and what action may be practical without overwhelming them with raw data.

The pilot produced 18 operational records and 14 completers. That experience tested device preparation, participant onboarding, communication, data continuity, and recovery from real-world interruptions. It established a practical foundation for evaluating the measure-to-action workflow, while also revealing where deployment procedures and system resilience need refinement.

The dissertation analysis combines operational records, exposure summaries, repeated self-reports, and user-experience evidence. The current phase evaluates field feasibility and prepares the planned sleep, mood, and productivity models for the manuscript in progress.