In-home sensor systems supported by machine learning are increasingly used to enhance communication between those living with long-term conditions such as dementia and healthcare professionals and carers who support them. Perspectives from the sociology of infrastructures are used to explore the development and deployment of such a system of smart care, drawing on interviews with researchers and developers, healthcare professionals and service users, and carers. The analysis finds that labor of various forms is required to manage the production of useful sensor data, including parsing the reasons for missing data and organizing appropriate actions in response. The analysis highlights active processes of deriving meaning from that data in ways that participants find useful, ethical, and sustainable. The conclusion emphasizes the usefulness of an infrastructural approach in order to recognize the heterogeneous forms of labor involved in developing ethically sensitive, person-centered forms of remote-monitoring-enabled care.

中文翻译：

---

协商家庭传感器介导护理的能力和局限性

由机器学习支持的家用传感器系统越来越多地用于增强患有痴呆症等长期疾病的人与支持他们的医疗保健专业人员和护理人员之间的沟通。通过对研究人员和开发人员、医疗保健专业人员和服务用户以及护理人员的采访，我们从基础设施社会学的角度来探索这种智能护理系统的开发和部署。分析发现，需要各种形式的劳动力来管理有用传感器数据的生成，包括解析丢失数据的原因并组织适当的响应行动。该分析强调了以参与者认为有用、道德和可持续的方式从数据中获取意义的积极过程。