Sensors and Data Analytics for Smart Homes

该项目的长期目标是设计能够监控居住者健康和福祉的智能家居。该项目专注于为老年人独立生活设计智能家居。让老年人在自己的家中保持更长时间的活跃可以提高他们的生活质量、认知能力和行动能力。此外，所提出的解决方案可以检测健康恶化的早期迹象，从而带来更好的医疗结果和健康老龄化。 本研究计划分为3个项目。第一个项目涉及环境中声音事件的检测和表征。该项目的目标是识别咳嗽、打鼾和尖叫等痛苦声音。将提出分类算法来监测呼吸状况并对紧急情况做出反应。第二个项目建议使用基于床的智能压敏垫解决方案来监测呼吸、离床特征和床上人员的体重分布。该项目的目标是及早发现一些健康状况，如睡眠呼吸暂停、睡眠质量问题、不稳定和充血性心力衰竭 (CHF)。在CHF中，体液白天积聚在下肢，晚上消散到身体的其他部位。第三个项目提出了新颖的上下文感知传感器融合解决方案，以最大限度地利用从传感器中提取的信息。深度学习算法将使用传感器模态替代，其中设计用于测量某种模态（例如温度）的传感器用于测量另一种模态（例如水流）。该项目的目标是提高系统性能、可靠性和成本。在每种情况下，该项目都将严格评估现有系统，评估可能的解决方案并推荐可行的解决方案。 这项工作将由研究生进行，他们将获得宝贵的多学科培训，并被鼓励在最负盛名的期刊和会议上发表他们的作品。 我们预计，除了上述学术和社会影响外，该项目还将具有巨大的经济发展潜力。拟议的解决方案将为加拿大公司在家庭和长期护理机构的健康和福祉监测领域创造新的商机。这包括寻求创建垂直市场的跨​​国电信公司、寻求新服务的基于云的数据分析公司以及寻求扩大业务范围的家庭安全公司。它还将为传感器和数据分析领域的初创公司创造新的创业机会。 为了确保我们提出的解决方案可行，我们的研究团队将定期与潜在用户（技术接受、需求评估、易用性、隐私）、医疗保健专业人员、政策制定者和工业设计师进行互动。

The long-term objective of this project is to design smart homes that are capable of monitoring the health and wellbeing of their occupants. The project focusses on the design of smart homes for the independent living of seniors. Keeping seniors active in their own homes longer improves their quality of life, cognition and mobility. Furthermore, the proposed solutions could detect early signs of health deterioration which could lead to better medical outcomes and healthy aging. This research program is divided into 3 projects. The first project deals with the detection and characterization of sound events in the environment. The objectives of this project is to recognize coughing, snoring and distress sounds such as screaming. Classification algorithms will be proposed to monitor respiratory conditions and react to emergencies. The second project proposes using a bed-based smart pressure-sensitive mat solution to monitor breathing, bed exit characteristics, and weight distribution of bed occupants. The objective of this project is the early detection of some health conditions such sleep apnea, sleep quality problems, instability and Congestive Heart Failure (CHF). In CHF body fluid accumulates in the lower limbs during the day and dissipates to the rest of the body at night. The third project proposes novel context-aware sensor fusion solutions to maximize the information extracted from the sensors. Deep learning algorithms will use sensor modality substitution where a sensor designed to measure a certain modality (e.g. temperature) is used to measure another modality (e.g. water flow). The objective of this project is to improve the system performance, reliability and cost. In each case, the project will critically assess existing systems, evaluates possible solutions and recommend feasible ones. The work will be conducted by graduate students who will gain valuable multidisciplinary training and will be encouraged to publish their work in the most prestigious journals and conferences. We expect that in addition to the academic and societal impacts described above, the project will have a significant economic development potential. The proposed solutions will create new business opportunities for Canadian companies in the area of health and wellbeing monitoring in homes and long-term care facilities. This includes multinational telecom companies looking for creating vertical markets, cloud-based data analytics companies looking for new services, and home security companies looking at expanding their scope. It will also create new entrepreneurship opportunities for start-up companies in the area of sensor and data analytics. In order to ensure that our proposed solutions are feasible, our research team will be interacting on a regular basis with potential users (technology acceptance, needs assessment, ease of use, privacy), healthcare professionals, policy makers and industrial designers.