Platform role
Home Assistant already provides the local device foundation.
Home Assistant runs on hardware in the home, stores its data locally, and can communicate directly with supported devices over the local network. Its core abstraction is the device and entity graph, with scenes, scripts, automations, dashboards, and integrations built around that state.
That foundation is valuable precisely because it should remain independently useful. HarborNavi's planned integration should consume only selected entities and allowed actions rather than turn Home Assistant into a hidden dependency that owners cannot inspect.
AI role
Home Assistant supports AI agents; the distinction is product scope.
Home Assistant can connect local or cloud language models and control the entities an owner exposes through its Assist and AI APIs. It also supports AI Tasks for work such as summarization and image analysis.
HarborNavi is being designed as a consumer household-intelligence product around that broader local context: selected camera events, device state, permissions, conversation, timelines, and temporary plans. The distinction is not that Home Assistant lacks AI; it is that the products organize the experience around different jobs.
Camera layer
Frigate and ONVIF solve different parts of the camera path.
Frigate is a local NVR designed around real-time object detection and Home Assistant integration. ONVIF is an interoperability framework, while RTSP is a media-control protocol. A household may use one or more of these layers depending on its cameras and desired events.
HarborNavi does not claim to replace every NVR or make every camera compatible. The planned starting point is selected RTSP and ONVIF paths, with support verified against real devices and network conditions.
- Keep recording and retention requirements explicit.
- Verify events, profiles, codecs, and credentials by model.
- Avoid duplicate camera connections when a restream can be used.
- Treat camera detection as one signal, not the whole household situation.
Fit
Choose the layer according to the job you want done.
Use Home Assistant directly when you want to build and maintain the automations yourself. Add Frigate when local camera recording and detection are central. Consider HarborNavi's product direction when the missing layer is household context, plain-language clarification, memory over time, and permission-aware coordination.
Because HarborNavi is still in development, households should evaluate it by supported integrations and tested workflows rather than assume compatibility from a logo or protocol name.