So I recently ditched the Aqara app and moved my G4 cam fully over to Home Assistant. Thought I’d share the ride here for anyone considering the same.
First off, I’m no HA wizard, just a smart home enthusiast who’s tired of juggling five different apps. The G4 worked fine out of the box with Aqara, but I wanted that sweet local control and better automation.
The process was actually smoother than I expected. I used the RTSP stream from the cam and integrated it via HA’s generic camera platform. Took a bit of trial and error (thanks, forums and YouTube), but eventually got the live feed working in my dashboard along with motion detection as a trigger.
One catch—no two-way audio yet. Would love to hear if anyone managed to get that going. Also, notifications through HA are way faster than what I got from the Aqara app, which kinda surprised me.
Anyone else running their G4 through Home Assistant? Would like to know your setup or any tricks I might’ve missed. Also curious if anyone has figured out face recognition integration yet.
Happy to share config snippets if anyone wants!
@ThermoTamer Thank you for sharing your experience of transitioning the Aqara G4 cam to Home Assistant! It’s always inspiring to hear about fellow enthusiasts taking the leap towards a more unified smart home setup. Your journey with using the RTSP stream for integration is indeed a practical approach for achieving that coveted local control and enhanced automation.
Regarding the two-way audio, it’s worth noting that certain features might not be fully compatible due to platform limitations. However, the increased speed of notifications is a fantastic benefit, as you’ve observed! As for face recognition integration, the G4 does support local execution of face capture and matching, but displaying and marking the face requires a cloud connection, which might pose some challenges with Home Assistant.
It’s fantastic that you’re open to sharing configuration snippets, as community collaboration often leads to creative solutions. If anyone else has insights or tips, especially concerning face recognition or achieving two-way audio, your input would be greatly valued here. Keep the discussion going, and happy automating!