PROJECT

Smart hydration sleeve

Weight and motion sensing for a familiar cup, developed within an interdisciplinary research team.

CONTRIBUTION

Co-development · Electronics integration · Initial sensing logic · Validation analysis

TOOLS & TECHNOLOGIES

ESP32-S3 · HX711 · IMU · C++ · MATLAB

STATUS

Working proof of concept · PCB integration in progress

Rand and Daniela Suarez beside their hydration sleeve research poster at Brown — Presenting the hydration sleeve with Daniela Suarez at Brown’s Summer Research Symposium. Open full photo ↗

Assembled hydration sleeve prototype from three angles — Team research prototype with sensing, display, and electronics integrated into the cup-holder platform. Open full image

01 / PROBLEM

Measure intake without changing the cup.

The research project explores hydration monitoring and reminders for older adults. A cup-holder platform combines a load cell, HX711, inertial sensing, an ESP32-S3, an OLED display, and an amplified speaker.

My contributions include electronics integration, initial C++ logic for stable weight-change tracking and motion rejection, the custom HX711 PCB, and MATLAB analysis of reference-scale comparisons. The broader reminder firmware and system were developed by the team.

02 / IMPLEMENTATION

Use motion to decide when weight is meaningful.

A moving cup produces weight readings that should not be interpreted directly as intake. The sensing approach suppresses measurements during movement, waits for a stable state, and compares the new mass with the previous stable measurement to estimate consumption or refilling.

For water, the system uses the approximation of 1 g per 1 mL. This is an intake estimate based on observed mass changes, rather than a direct measurement of swallowing.

Breadboard and cup bench setup — Early physical testing environment from my project showcase. Open full image

03 / PROBLEM SOLVING

Consider the causes. Start with the simplest fix.

Comparing the prototype with a laboratory reference scale revealed a drifting cup-mass baseline and a mismatch in per-sip estimates. I analyzed the discrepancies in MATLAB; the two poster graphs below show this early behavior.

We considered several possible sources of the discrepancy. Improving the cup-to-sleeve coupling was the first and simplest intervention to try. It resolved the observed problem: the gap between device and reference measurements disappeared after the fix.

Other possible fixes remained part of our troubleshooting plan if the coupling change had not worked; they were not needed to resolve this issue in the proof of concept.

Research-poster graph comparing device-tracked cup mass against the reference scale across sequential trials, showing a drifting baseline — Baseline drift before the fix. The gap shown here disappeared after improving the cup-to-sleeve coupling. Open graph ↗

Research-poster graph comparing device-tracked per-sip hydration with the gravimetric reference — **Per-sip estimates.** The poster’s intake comparison from the same early prototype. Open graph

04 / PCB DEVELOPMENT

Turn the proof of concept into a compact platform.

The development-board prototype demonstrated weight and motion sensing together with display and audio feedback. The custom PCB builds on that working concept: replace separate boards and jumper wiring with a smaller, more integrated assembly that fits naturally into the sleeve.

So far, I have designed the HX711 front end and its PCB layout. Development of the rest of the integrated board is in progress.

← All projects Discuss this project Next projectEMG-controlled claw