Collaborative Research: SCH: Improving Older Adults' Mobility and Gait Ability in Real-World Ambulation with a Smart Robotic Ankle-Foot Orthosis

该项目将开发一种新型外骨骼，以改善在现实环境中的行走。当人们变老，肌肉功能下降时，就会限制他们产生足够的踝关节推力来正常行走的能力。行动不便的老年人经常走路不自然，效率低下。例如，他们可能会通过过度使用臀部来弥补踝关节推离动作的减少。这种行为不仅影响相关腿部关节的长期健康，还会导致身体活动能力加速下降。这种活动能力的下降对老年人的生活方式以及他们长期的身心健康产生了深远的影响。随着步行速度的减慢，长者的日常活动可能会遇到意想不到的困难（例如，在交通灯变亮前无法穿过繁忙的街道）。当步行变得更加费力时，老年人更有可能缺乏运动，并遭受与这种久坐不动的生活方式相关的多种健康问题（高血压、肥胖、抑郁等）。受到这一重大挑战的激励，我们在这个项目中的研究将致力于开发一种新的机器人踝足矫形器（外骨骼）系统，以帮助用户在现实环境中行走时的脚踝运动。在机器人矫形器的辅助下，老年人可以更自然、更高效地行走，增强踝关节蹬蹬，从而在日常生活中显著提高步态能力和身体活动能力。该项目将开展多项研究活动，以创建拟议的机器人踝足矫形器（AFO）系统。它将设计和制造一种紧凑轻便的日常使用机器人踝足矫形器（duro - afo），为用户提供所需的物理辅助，并且几乎没有额外的负荷。通过对脚踝的动力辅助，dura - afo有望诱导老年人增加他们的脚踝蹬蹬，以获得更自然和有效的步态（更接近年轻健康的步态），从而提高他们的步态能力。此外，该项目将开展生物力学研究，研究在独立控制的双边机器人辅助下，人类在现实世界运动（行走和转弯）中的步态控制机制；我们还将探索使用实时信息反馈（振动触觉提示，音频提示等）的新方法，以授权和激励用户在享受dura - afo的动力辅助的同时保持所需的肌肉努力水平。最后，该项目将开发一种新的基于强化学习（RL）的网络系统，作为人机协同协作系统的基础，提供多种重要功能，如适应机器人控制参数以进行步态质量优化，确定对人类用户的实时反馈，以及以期望的运动模式的形式识别人类运动意图。总的来说，这种新型的人机协同协作框架不仅优化了机器人辅助人类运动的性能，而且还促进了人类用户有益的行为改变（“在步行锻炼中保持所需的肌肉努力水平”），以实现人机系统的共同目标（“提高人类在日常生活活动中的步态能力和整体机动性”）。该项目由智能和互联健康以及刺激竞争研究的既定计划（EPSCoR）共同资助。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This project will develop a novel exoskeleton to improve walking in real-world environments. When people age and their muscle functions decline, it can limit their ability to generate enough ankle push-off power to walk normally. Mobility-challenged older adults often walk unnaturally and inefficiently. For example, they may compensate for reduced ankle push-off action by overusing the hip. This behavior not only affects the long-term health of the related leg joints, but also causes accelerated decline of physical mobility. Such mobility decline has profound impacts on older adults’ lifestyles, as well as their long-term physical and psychological health. As walking speed decreases, an elderly individual may experience unexpected difficulties in his/her daily activities (e.g., being unable to cross a busy street before the light changes). When walking becomes more strenuous, an older adult is more likely to be physically inactive and suffer from the multiple health problems associated with such a sedentary lifestyle (high blood pressure, obesity, depression, etc.). Motivated by this significant challenge, our research in this project will be dedicated to the development of a new robotic ankle-foot orthosis (exoskeleton) system to assist the user’s ankle movement when walking in real-world environments. With the robotic orthosis’ assistance, older adults may walk more naturally and efficiently with the enhanced ankle push-off, and thus enjoy significantly improved gait ability and physical mobility in their daily lives.The project will conduct multiple research activities towards the creation of the proposed robotic ankle-foot orthosis (AFO) system. It will design and fabricate a compact and lightweight Daily-Use Robotic Ankle-Foot Orthosis (DUR-AFO) to provide the desired physical assistance with little additional load to the user. Through its powered assistance to the ankle, the DUR-AFO is anticipated to induce elderly individuals to augment their ankle push-off for a more natural and efficient gait (closer to younger healthy gait) and thus improve their gait ability. Further, the project will conduct biomechanical research to investigate the human gait-control mechanisms under independently controlled bilateral robot assistance in real-world locomotion (walking and turning); we will also explore the novel approach of using real-time information feedback (vibrotactile prompts, audio cues, etc.) to empower and motivate users to maintain a desired level of muscle efforts while enjoying the powered assistance by the DUR-AFO. Finally, the project will develop a novel Reinforcement Learning (RL)-based cyber system as the basis of the human-robot synergistic collaborative system, providing multiple important functions such as adapting the robot control parameters for gait quality optimization, determining the real-time feedback to the human user, and identifying the human motion intent in the form of the desired mode of locomotion. Overall, this novel human-robot synergistic collaboration framework not only optimizes the performance of the robot assistance to the human movement, but also promotes the human user’s beneficial behavioral changes (“maintaining a desired level of muscle efforts in walking exercise”) towards the shared goal of the human-robot system (“improving the human’s gait ability and overall mobility in daily-life activities”). This project is jointly funded by Smart and Connected Health and the Established Program to Stimulate Competitive Research (EPSCoR).This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.