Towards optimization of wearable sensor technology to measure human movement in real world settings

近年来，用于测量和跟踪人体功能的可穿戴传感器市场呈爆炸式增长。越来越多的人使用这种设备来监控日常生活的许多不同方面。这种爆炸性增长是由技术进步带来的，例如这种传感器的尺寸和成本降低，以及电池寿命和存储能力的增加，并且由于公众越来越希望更好地了解人体如何工作而促进了这种爆炸性增长。不幸的是，这些设备的可用性和需求的增长远远超过了它们所创建的数据科学。存在非常少的信息来告知与最佳传感器配置、某些变量的适当性以及最佳测量的活动相关的决策。拟议的研究将直接解决这些差距，并将导致一种新的分类算法，旨在使用可穿戴技术检测真实的世界中运动参数的变化。这些都是必要的步骤，以实现我的长期愿景，即通过开发创新的检测算法和最佳实践策略，利用可穿戴技术收集运动数据，更好地了解基本的人体运动。该研究的短期目标是在日常生活的常见活动中使用可穿戴传感器技术来改善人类运动模式的量化。在接下来的五年里，我将领导一个研究项目，这将提高我们在真实的世界环境中理解人类运动的能力。我们将使用复杂的分析技术改进传感器配置和与从可穿戴传感器收集人体运动数据相关的关键参数。我们的假设和进展将首先使用当前的黄金标准运动捕捉系统进行通知和测试，以确保我们的新传感器配置的准确性和有效性。然后，我们将通过开发一种新的检测算法，旨在对现实世界中的运动变化进行分类，将我们的发现转化为真实的世界。** 我们从这项研究中获得的新贡献将为研究人员提供更好的知识，以了解如何最好地将可穿戴运动技术纳入他们的研究中。这些知识将有助于促进从依赖昂贵的基于实验室的运动评估的过渡，这种评估可能无法准确地描述真实的世界功能。工业界将利用这项研究的成果来完善新的和现有的商用可穿戴运动传感器的开发，从而为普通人群提供更准确、更可靠的产品。最后，公众将从这项工作中受益。从使用这些设备来评估运动以提高运动表现的运动员和教练到对简单地跟踪其运动表现以获得整体健康感兴趣的日常个人的用户，都将受益。

The market for wearable sensors to measure and track human function has exploded in recent years. More and more people are using such devices to monitor many different aspects of everyday life. This explosion has been brought on by advances in technology such as reduced size and cost of such sensors, as well as increased battery life and storage capabilities, and has been facilitated by an ever-growing public desire to better understand how one's body is working. Unfortunately, this growth in availability and demand for these devices has far outgrown the science of the data in which they create. Very little information exists to inform decisions related to best sensor configuration, appropriateness of certain variables, and which activities are best measured. The proposed research will directly address these gaps, and will lead to a novel classification algorithm designed to detect changes in movement parameters in the real world using wearable technology. These are necessary steps to achieve my long-term vision of a better understanding of fundamental human movement through the development of innovative detection algorithms and best practice policies for the collection of movement data using wearable technologies.******The immediate short-term objectives of the research will be to improve the quantification of human movement patterns using wearable sensor technology during common activities of daily living. I will lead a research program over the next five years that will advance our ability to understand human movement in real world settings. We will improve sensor configuration and key parameters related to the collection of human movement data from wearable sensors using sophisticated analysis techniques. Our hypotheses and advances will first be informed and tested using current gold-standard motion capture systems to ensure accuracy and validity of our new sensor configurations. We will then translate our findings to the real world by developing a novel detection algorithm aiming to classify movement changes in real-world settings. ******Our novel contributions from this research will provide researchers with better knowledge for how to best incorporate wearable movement technology into their studies. This knowledge will help facilitate a transition from reliance on expensive laboratory-based movement assessment that may not accurately depict real world functioning. Industry will use the contributions from this research to refine the development of new and existing commercially-available wearable movement sensors that will result in more accurate and reliable products used in the general population. Finally, the general public will benefit from this work. Users ranging from athletes and coaches who use these devices to assess movement to improve athletic performance, to everyday individuals who are interested in simply tracking their movement performance for overall well-being, will stand to benefit.