Shower room is essential in our life that people are spending significant amount of time in a day, e.g., washing hand at sink, flushing toilet bowl, taking shower. This is especially significant at South East Asia region, in which people often take shower more than one time in a day because of the hot & humid weather. This means significant amount of water went to the drain every household every day, without noticing how much water was wasted because people usually just left the shower head running continuously throughout the shower, in fact we only need the water when rinsing our body. There are various study available online to indicate the amount of water we used during shower, but none of it really get to you & me because no one really pay attention to the water footprint.
The idea is to have Nordic Thingy:53 to run tinyML for determining if it's a hand-washing event, shower, or flushing and record the duration. The analysis later on can tell what is our average value and then setting goal which to define a threshold for each shower, switching on the buzzer when the showering event is longer than expected. However, the current work is only partially achieved with the inferencing part, the later part of recording and triggering is not done yet because of the learning curve of nRF Connect SDK, I find it difficult to navigate through Zephyr OS. I do hope that I have more time to digest the detail of Zephyr & work with nRF Connect SDK.
Comments