Presented at MATLAB EXPO 2026 and Took 2nd Place in the Poster Session
We presented the development and field validation of a surface observation system for drift prediction in Lake Biwa, and took 2nd place in the poster session.
At the MATLAB EXPO 2026 poster session, AquaWiz's Kamiyama presented a surface observation system for drift prediction in Lake Biwa and took 2nd place. Kamiyama led the research as first author, working with two high school students outside AquaWiz as a three-person team.
The research aims to collect observation data that can help clarify how drifting objects move in Lake Biwa, supporting uses such as disaster rescue and predictions of drifting algae washing ashore. While observation in Lake Biwa is centered mainly on fixed-point measurements, data from drifting observation devices remains limited. To address that gap, the team developed a compact observation system that can drift across the lake surface while recording position, water temperature, and surrounding environmental data.
The team also developed a monitoring web app that lets operators check data acquired over LTE from shore. The app shows the current position of the device and its observation data, and includes features such as position prediction and data export to support operation during release experiments.
In a long-duration release experiment in northern Lake Biwa, the team confirmed issues including device malfunction and water ingress. At the same time, by combining GPS logs with nearby weather data, the experiment provided clues for considering surface flow and the influence of local winds. The team also worked on battery state-of-charge estimation using a coulomb counter for the power management needed in long-duration observation.
The presentation introduced not only the development of the observation device, but also the integrated work on communication, visualization, power management, and field validation.