Peace of Mind at Home: A Supportive Home Assistant Setup for Elderly Loved Ones

Peace of Mind at Home: A Supportive Home Assistant Setup for Elderly Loved Ones
Photo by Mr Cup / Fabien Barral / Unsplash

Following my grandmother's recent fall and subsequent stay in rehab, I sought a way to enhance her monitoring and communication, aiming to preserve her independence and ensure her ongoing well-being. I leveraged Home Assistant, a system I am very familiar with, as a suitable platform for this task.

It's important to stress that while Home Assistant is reliable, I do not use it for any critical medical or safety-related functions. This setup is purely for augmentation; its failure would be an inconvenience, but would not compromise her health or safety. This setup has been performing successfully for about a month now.

The Hardware

I chose a mix of reliable, cost-effective hardware suitable for a small one-bedroom apartment:

  • Compute: A Raspberry Pi 5 (8GB) was chosen for its small form factor, simple setup, and good value.
  • Z-Wave Connectivity: I used a Zooz Z-Wave 500 Stick and two Aeotec Range Extender 7s to ensure a strong mesh network.
  • Presence and Motion Sensing: * 4 x Ring Gen2 Motion Sensors: These use common AA batteries and are cost-effective.
    • 1 x Ring Open/Close Sensor: Used for monitoring doors and windows.
    • 1 x Aqara FP2 mmWave Sensor: Provides high-accuracy room presence and zoning capabilities.
  • Smart Lighting & Plugs: A mix of Inovelli Gen 1 Dimmers and smart plugs from TP-Link Kasa and other reliable brands.
  • User Input: A GoveeLife Smart Mini Double Button Switch (BLE) for simple physical control.
  • Safety (Independent): A QMedic emergency bracelet with cellular backhaul and backup battery. By design, this is not tied to Home Assistant to ensure it works during power or phone outages.

Use Cases

The goal of this setup is to provide subtle but effective support:

  • Automated Lighting: Lights turn on at sunset automatically so she never ends up in the dark.
  • Bedside Control: A button right at the bedside turns off smart plugs and lights, allowing her to transit the home with proper illumination before getting into bed.
  • Bed Presence Detection: The mmWave sensor determines if she is in bed versus just near the bed.
  • Activity Tracking: Motion sensors and door sensors track recent activity and the last time anyone entered or exited the apartment.
  • Emergency Communication: I can use two-way audio to communicate if the phone is off the hook or if she is unable to get to it.

Installation & Software

Home Assistant Installation

Installing Home Assistant on the Raspberry Pi 5 is easily managed using the Pi Flasher application. You simply prepare the storage, select “Home Assistant” in the wizard, and flash the image. Once flashed, you insert the card into the Pi, power it on, and it automatically sets up the environment.

Essential Add-ons & Integrations

  • Cloudflared: Securely exposes Home Assistant to the internet via a tunnel.
  • Z-Wave JS UI: Manages the mesh network for all sensors.
  • Home Assistant Time Machine: Keeps track of automation changes for easy rollbacks with clear diffs.
  • Studio Code Server: For editing configuration files when required.
  • HACS: Provides integrations like WebRTC Camera and Local conditional card.

Camera Setup

The monitoring setup uses two Tapo C110 and one C210 camera. These are inexpensive, support RTSP streams, and offer two-way audio.

To get these working, you first connect them to the Tapo app, enable RTSP streams, and generate a SHA256 hash of your cloud password. I’ve documented this in my post about getting the Tapo cameras working with Frigate.

Go2rtc & WebRTC

I use the go2rtc add-on to connect to the Tapo cameras using a proprietary protocol for two-way audio. It also centralizes and optimizes streaming to take the load off the cameras.

Your streams config may look something like this. You'll note, added a line for reencoding the audio. This is to ensure broad compatibility when streamed from various devices.

streams:
  living:
    - tapo://admin:[email protected]
    - ffmpeg:living#audio=aac#audio=opus
  bedroom:
    - tapo://admin:[email protected]
    - ffmpeg:bedroom#audio=aac#audio=opus
  kitchen:
    - ffmpeg:kitchen#audio=aac#audio=opus
    - tapo://admin:[email protected]

However, because this setup is behind a double NAT, I couldn't initially use WebRTC for peer-to-peer communication. To traverse this networking hurdle, I utilized a free tier from Open Relay, which provides 20GB of monthly streaming—plenty for this use case. This enables reliable, low-latency video and functional two-way audio directly from the Home Assistant dashboard.

Your webrtc: section may look something like this:

webrtc:
  ice_servers:
    - urls:
        - stun:stun.relay.metered.ca
    - urls:
        - turns:standard.relay.metered.ca:443?transport=tcp
      username: "replace_with_username"
      credential: "replace_with_password"

Conclusion

Building this Home Assistant environment has provided a meaningful layer of support for my grandmother, successfully augmenting her daily routine while preserving her independence. By combining high-accuracy mmWave sensing for bed presence with reliable automated lighting and remote communication tools, the system addresses specific environmental risks—like navigating in the dark—while balancing privacy implications.

Ultimately, this setup demonstrates how home lab enthusiasts can leverage their existing skills to provide practical, non-critical care solutions for loved ones. While it is never a replacement for professional medical services or dedicated emergency hardware like her cellular bracelet, it offers a significant “peace of mind” bridge that makes aging in place just a little bit safer.