Smart Home Presence Detection: Beyond Motion Sensors

The single biggest frustration in smart home automation is lights turning off while you are still in the room. You are sitting on the couch reading, the PIR motion sensor has not detected movement for 5 minutes, and suddenly you are in the dark waving your arms like you are directing air traffic. Sound familiar?

The problem is not the automation, it is the sensor. PIR (passive infrared) motion sensors detect movement, not presence. If you stop moving, you disappear. That is a fundamental limitation of the technology, not a configuration issue. No amount of timeout tweaking fixes it completely.

Let me walk you through the alternatives that actually solve this problem.

mmWave Radar: The Game Changer

Millimeter-wave radar sensors detect the micro-movements of a human body, breathing, heartbeat, slight shifts in posture. You can sit perfectly still reading a book and an mmWave sensor knows you are there. This is the technology that has genuinely solved presence detection for smart homes.

The most popular option is the HiLink LD2410 module, which costs about $5-8 and integrates with ESPHome. Pair it with an ESP32 and you have a room presence sensor for under $12. The LD2410 has configurable detection zones (close range, medium range, far range) and sensitivity thresholds that let you tune it for each room.

Placement Tips for mmWave

• Mount it high. Ceiling or high-wall mount gives the best coverage angle. A sensor looking down into a room has fewer blind spots than one at desk height.
• Avoid pointing at fans or curtains. mmWave detects any movement, including ceiling fans and curtains blowing in the breeze. If you cannot avoid it, adjust the sensitivity zones.
• One sensor per room. Unlike PIR, a single mmWave sensor can cover a standard room (up to about 20 square meters) from a central position.
• Keep it away from glass. Radar penetrates thin walls and glass. A sensor near a window might detect people outside or in the next room, causing false positives.

Bluetooth Room Detection

If you carry a phone or wear a smartwatch, Bluetooth-based detection can tell which room you are in. Home Assistant supports this through ESPresense, ESP32 boards that listen for Bluetooth advertisements from your devices and triangulate your position based on signal strength.

The setup requires at least one ESP32 in every room you want to track. Each board runs ESPresense firmware and reports the signal strength of your phone or watch. Home Assistant uses these readings to determine which room you are closest to.

The accuracy is not perfect, Bluetooth signal strength varies with phone orientation, body position, and obstacles. But for room-level granularity (kitchen vs living room vs bedroom), it works well enough for automation triggers.

WiFi-Based Detection

Some routers and access points can detect presence based on WiFi signal disruption. When a person moves through a room, their body affects WiFi signals between the router and connected devices. Companies like Linksys (with Aware) and some UniFi setups can leverage this for basic occupancy detection.

The accuracy is limited compared to mmWave, and it typically only provides house-level presence (home vs away) rather than room-level. It is best used as a supplementary signal rather than a primary trigger.

Multi-Sensor Fusion: The Best Approach

The most reliable presence detection combines multiple sensor types. Here is a stack that covers the gaps:

The logic in Home Assistant: PIR triggers instant light-on (fast response). mmWave sustains the light while someone is present (no false-offs). Door sensor helps determine entry and exit events. Bluetooth identifies who is there for personalized scenes.

Is it overkill? For a bathroom, yes. For your living room where you spend hours sitting still, it is the difference between a smart home that works and one that annoys you daily.