Dehydration poses a significant health risk for older adults, especially during heatwaves, yet daily fluid intake is rarely monitored. This paper presents the design of a wearable sensor–based framework that leverages inertial data from off-the-shelf smartwatches to recognise drinking gestures and provide hydration feedback through a companion web application. Following a user-centred design process, we evaluated early interface prototypes with 30 carers and relatives of elderly people in Greece and Mexico using the System Usability Scale (SUS). Results placed the prototype at the upper end of the “good” usability band (M = 77.8, α = 0.91), with half of respondents rating it “best imaginable.” User feedback highlighted design requirements such as vibration-based reminders, dual-hand tracking, and caregiver dashboards. While these results demonstrate strong usability perceptions, the study did not include elderly end-users and did not yet validate the technical accuracy of gesture recognition. Future work will therefore involve controlled trials with older adults, motion-detection validation against confounding activities, and integration of fluid-intake estimation to more directly assess dehydration risk.