Multisensor Systems for Physical Activity Monitoring and Assessment

点击翻译按钮获取中文摘要

Most Canadians are not realizing the benefits of regular physical activity, including increased energy, better relaxation, increased strength and stamina, improved weight management, and, most importantly, a reduced risk of many chronic conditions and diseases. In the absence of an expert or a coach, a persistent challenge for many engaging in activity is knowing whether they are doing things correctly-gaining the benefits of higher intensity exercise while minimizing the risk of injury. For this purpose, the proposed research program is focused on solving the engineering challenges associated with the development of accessible technologies to guide and support individuals in their physical fitness activities. To address these challenges, the short-term objectives of the proposed research are as follows: 1. To correctly identify specific physical activities that individuals are doing based on real-time images and data from wearable sensors. Machine learning methods will be developed to robustly discriminate exercises in real time. 2. To develop metrics that can accurately estimate the energy used during physical activities from a variety of sensors. The optimal set of sensors will be determined and new models for energy expenditure estimation will be developed. 3. To develop techniques and metrics that can assess and quantify how closely activities performed by an individual compare to how they should be done. Motion models will be developed along with metrics that indicate motion quality at the joint level. 4. To design and evaluate the best haptics-based approaches to provide feedback to an individual during physical activity. A variety of actuator configurations and stimulation patterns will be explored to determine the best way to deliver haptic cues for exercise guidance. The outcomes of the proposed research over the next 5 years will lead to significant contributions in the development of systems and methods that can interpret visual and wearable sensor data to provide insights and actionable feedback during physical activity. Together, they will provide additional tools to support broader initiatives to increase activity levels. In the long term, the proposed research will result in accessible technologies that help people engage in physical activity safely and effectively without the presence of a health expert, trainer, or coach. To succeed with the proposed work, equity, diversity, and inclusion criteria will be considered when hiring research personnel, when recruiting participants, when analyzing the data, and in the development and translation of the technology. A collaborative approach will be taken to consider the needs of the users and other key stakeholders.