01Make common actions easier
Assess clearly labelled controls for lighting, curtains and supported climate settings. A bedside scene or a single room button may be more useful than a complex app. Try control locations and interaction methods with the resident, including a way to cancel an unwanted action. Voice can be optional where it suits the person and the equipment.
02Agree consent before adding observation
Explain what a proposed sensor records, who can see it, where information is stored and how long it is retained. Monitoring should never be hidden from the resident. Consider whether a less intrusive control or alert meets the need. Document the agreed permissions and an understandable way to pause sharing or revisit the arrangement.
03Define what an alert actually means
A door event or a period without sensor activity does not establish someone’s wellbeing. If a supported notification is included, agree the event, recipient, expected response and a backup contact arrangement. Test delivery with those people. Notifications may be delayed, missed or triggered incorrectly; neither a delivered alert nor silence confirms that someone is safe.
04Keep ordinary care and emergency arrangements
Consumer home automation is not medical monitoring, a fall-detection guarantee or an emergency-response service. It does not replace carers, appropriate clinical advice or existing emergency arrangements. Google’s automation guidance, for example, warns that routines depend on connectivity and service availability. Plan manual alternatives and periodic checks without relying on an automation for a critical care task.
The proposed scope is limited to compatible home controls and agreed convenience notifications. We do not promise continuous observation, clinical outcomes or emergency response.