MLH Hack At Home 2020: Hearth
This is a submission for the MLH x DEV Writing Challenge What I Built Hearth is a voice-controlled smart home agent built for the moment we were all in: stuck at home, working from the same room we slept in, and slowly losing track of what day it was. The theme for Hack at Home was "help folks stay happy at home." In May 2020, the problem in my apartment wasn't a lack of smart devices. It was…
MLH Hack At Home 2020: Hearth is a voice-controlled smart home agent designed to help people navigate the challenges of living at home during the COVID-19 pandemic. Created in May 2020, Hearth addresses the issue of smart devices being integrated into every aspect of daily life, with no clear separation between work, gym, classroom, and bedroom. The project achieves three main objectives:
1. Voice control that accurately identifies devices: Hearth leverages Home Assistant's real device names and integrates them into its language model, resulting in more precise command execution compared to generic commands.
2. Modes for a cohesive living experience: Hearth offers mode-specific settings for various activities, such as work mode, which adjusts lighting, temperature, and notifications to create an optimal environment for productivity.
3. Pattern suggestions from sensor history: Hearth analyzes historical data from Home Assistant to propose automations that users might perform manually, such as turning on kitchen lights at a specific time each weekday.
The project utilizes Picovoice Porcupine for offline wake word detection, Google Cloud Speech-to-Text for transcribing commands, Dialogflow ES for intent understanding, and a Python Flask webhook on Google Cloud Run for fulfilling intents. Safety features, such as requiring approval for locking and garage actions, are implemented to ensure user control over critical functions.
The Android app, developed using Kotlin and coroutines, provides an interface for managing devices, pending approvals, and suggested automations. The Hearth smart home agent was built for the MLH x DEV Writing Challenge, and the source code is available on GitHub.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
