snazzybean/roommind

RoomMind is a custom integration for Home Assistant (the popular open-source home automation platform) that brings intelligent, self-learning climate control to individual rooms in your home. The problem it solves is that standard smart thermostats and radiator valves use simple on/off logic that causes temperature swings and wastes energy, they don't learn how your specific rooms heat and cool over time. The way it works is through a thermal model that watches how each room's temperature actually responds to heating, cooling, and sunlight, then uses that learned behavior to make smarter decisions. It calculates proportional setpoints for radiator valves and air conditioners, meaning instead of just switching devices on or off, it sends precise target values to hit your desired temperature smoothly. It also accounts for solar gain (how much sunlight is warming a room), presence detection (setting lower "eco" temperatures when everyone's away), window and door sensors that pause heating, mold risk prevention, and automatic blind control to block overheating from sun exposure. You would use this if you have Home Assistant set up and want room-by-room climate control that actually learns your home's thermal behavior rather than relying on simple schedules or basic on/off switching. It installs as a custom component through HACS (the Home Assistant community store) and adds a management panel in the sidebar with analytics, scheduling, and per-room controls. The code is written in Python.