SESAME: Sensing for Sport and Managed Exercise

SESAME联盟是一个新成立的多学科小组，旨在研究使用无线传感器系统来提高被认为具有世界级潜力的精英运动员和年轻运动员的表现。该项目的目标是提高成绩，改善教练教育，推进体育科学。尽管特别关注体育运动，但技术方法及其解决方案将故意通用，以使其后续应用于更广泛的培训和医疗场景。目前，只有一套有限的传感技术可用于精英运动员的训练，包括动作捕捉，固定力板和视频记录反馈。然而，他们经常扰乱体育活动，他们返回的数据很难解释，以提供适当的反馈。无线传感技术，从加速度计、磁力计到精确定位系统，通过实时提供肢体定位和方向、运动员位置、肌肉功能和生理状态等信息，有能力彻底改变这一领域。通过SESAME项目，动态数据将来自可穿戴式非侵入式传感器，并通过被动视频捕获进行增强。原始传感器数据将被处理以提取有意义的信息，使用传感器融合和随机信号处理的组合来获得对教练和运动员有意义的信息。这将发生在人类生物力学限制运动的知识中，并将考虑由传感器附着机制和传感器错误定位引入的误差。捕获的传感器数据将为生物力学和生理性能模型提供信息，并从中捕获理想运动和性能偏差的影响。如果教练和运动员要从SESAME中获得真正的好处，对人为因素进行全面的研究是必不可少的。将在教练的指导下进行人种学研究，以建立特定领域的专家知识，掌握他们的表现认知模式，并协助设计用户界面。对教练和运动员的反馈将以两种形式呈现：(i)图形化，这两种形式都是经过处理的数据流，以尊重教练的认知模型，并将传感器数据叠加在视频上；（ii）如可行，作为实时反馈：例如使用蜂鸣器。对运动员表现的分析不仅是一项实时活动：传感器数据、决策支持建议、医疗建议和任何临床事件的明确记录将得到保存，使用户能够考虑相关的医疗投入。这种方法还允许运动员之间的比较研究，并挖掘这些信息，以改进生物性能模型，并了解偏离理想状态和损伤前兆的影响。工作的重点将放在跑步上，特别是短跑。然而，考虑到2012年奥运会对国家的重要性，我们也将探索在其他运动项目、更一般的运动形式和受伤后康复中使用这项技术的可能性。如果时间允许，更广泛的应用，如步态分析脑瘫患者也将探索。从无线传感器网络的角度来看，运动训练是一个要求很高的应用领域：有必要开发和集成新型传感器、qos驱动的实时网络和系统自动配置，所有这些都使用可扩展的通用软件基础设施。因此，在这个具有挑战性的领域中解决问题将为解决无处不在和感知计算中更普遍的问题提供必要的构建块。SESAME联盟包含了对在这一领域部署技术取得进展至关重要的专业知识。

The SESAME consortium is a newly-formed multidisciplinary group that proposes to investigate the use of wireless sensor-based systems in enhancing the performance of elite athletes and young athletes who have been identified as having world class potential. The project has goals of enhancing performance, improving coach education, and advancing sports science. Despite a specific focus on athletics, the technical approach and its solutions will be deliberately generic, to enable their subsequent application to a wider range of training and healthcare scenarios. At present, only a limited set of sensing technologies are available for the coaching of elite athletes, including motion capture, fixed force plates and video recording for feedback. However, they often disrupt the sporting activity and the data they return are difficult to interpret to provide appropriate feedback. Wireless sensing technologies, ranging from accelerometry and magnetometry through to accurate positioning systems, have the capacity to revolutionise the field, by providing information about limb positioning and orientation, athlete location, muscular function, and physiological status, all in real time. Through the SESAME project, dynamic data will come from wearable non-intrusive sensors, augmented by passive video capture. Raw sensor data will be processed to extract meaningful information using a combination of sensor fusion and stochastic signal processing to derive information that is meaningful to coaches and athletes. This will take place in the knowledge that human biomechanics constrains movement and will take account of errors introduced by sensor attachment mechanisms and sensor mispositioning. Biomechanical and physiological performance models will be informed by captured sensor data, and from them idealised movements and the performance effects of deviations will be captured. A comprehensive study of human factors is essential if coaches and athletes are to derive real benefit from SESAME. Ethnographic studies will be undertaken with coaches - to build expert domain-specific knowledge, to capture their cognitive models of performance, and to assist in the design of user interfaces. Feedback to coaches and athletes will be in two forms: (i) graphical, both as a data stream that has been processed to respect the coaches' cognitive models and by overlaying sensor data on video; (ii) as real-time feedback if feasible: e.g. using buzzers. Analysis of an athlete's performance is not only a real-time activity: a definitive record of sensor data, decision support recommendations, medical advice and any clinical events will be maintained, allowing users to take account of relevant medical inputs. Such an approach also allows for comparative studies between athletes and the mining of such information both to improve biological performance models and to understand the effect of deviation from the ideal and precursors to injury. The focus of the work will be on running - specifically sprinting. However, given the national importance of the 2012 Olympic Games we will also explore the possibility of using the technology in other athletic disciplines, more general forms of exercise, and rehabilitation following injury. Should time permit, wider applications such as gait analysis for cerebral palsy patients will also be explored. Athletic training is a highly demanding application domain from the viewpoint of wireless sensor networking: it is necessary to develop and integrate novel sensors, QoS-driven real-time networking, and system autoconfiguration, all using an extensible generic software infrastructure. Consequently, solving problems in this challenging domain will provide a necessary building block for the solution of more generic problems in ubiquitous and sentient computing.The SESAME consortium contains a blend of expertise that is essential for progress in deploying technology in this domain.