CAREER: Dynamic Locomotion with Plasticity for Remote Sensing in Crawlspaces

爬行空间无处不在：洞穴、喀斯特地貌、竖井、管道、压载舱、管道和谷物升降机。来自这些环境的数据具有巨大的科学意义（生物多样性和地质过程）、社会意义（考古和文化遗迹）、环境意义（自然资源和矿藏）以及健康安全经济意义（资产管理和监测）。然而，在爬行空间中人工操作成本高、速度慢且有风险。现有的蛇和昆虫机器人可以在管道中爬行，也可以在厘米级的空间内操作；然而，它们速度太慢，或者不是完全自主的。我们迫切需要专门为爬行空间设计的快速、敏捷和高效的自主系统。腿式运动和多旋翼飞行有望解决这一挑战，已经通过超越任何其他机器人类别的能力彻底改变了许多遥感任务。然而，他们的行动仍然局限于宽敞的隧道。这个学院早期职业发展（Career）项目支持的研究旨在利用腿式和旋翼机器人的卓越敏捷性、效率和速度，并使其专门用于爬行空间。该项目设计了一种具有广泛运动可塑性的多模态机器人，能够通过双足行走和飞行，高效地穿越极其狭窄的爬行空间。此外，该项目旨在促进东北大学机器人项目中的性别平等，这是学校领导层确定的一个重要增长领域。在爬行空间中实现高效、敏捷和快速的腿式空中运动是机器人运动的一个新进展。在爬行空间中，桥接腿式机器人和多旋翼飞行仍然存在三大知识空白：(1)驱动挑战阻碍了运动控制性能的可扩展性，这是在开放空间中设计的腿式机器人向在爬行空间中运行的小型机器人过渡时快速精确地放置足部所必需的；(2)多旋翼近地面喷气产生的不稳定问题存在飞行固定风险；(3)爬行空间需要多种运动模式，没有系统的机器人设计框架来适应这些模式的要求。该项目将进行基础研究，以解决这些空白，设计一种仿生运动机器人，能够在倾斜表面上行走、悬停、跳跃和奔跑，以推动移动机器人的操作范围，使在极其狭窄的爬行空间内自主操作成为可能。该项目的工作围绕着机器人和控制设计。该项目的关键创新领域包括：(1)基于计算结构设计引入小型机器人的驱动设计范式，以实现大型机器人的运动控制性能；(2)基于姿态操纵与推力矢量集成的新型运动技术及其基础模型和非线性控制器的研究与验证，以克服航空机器人中的喷气风险；(3)通过生成式设计方法对机器人和控制器进行协同设计，以适应多种运动模式带来的冲突要求。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。

Crawlspaces are everywhere: caves, karst landscapes, shafts, ducts, ballast tanks, pipes, and grain elevators. Data from these environments holds immense scientific (biodiversity and geological processes), societal (archaeological and cultural remains), environmental (natural resources and mineral deposits), and health-security-economical (asset management and monitoring) significance. However, human operation in crawlspaces is costly, slow, and risky. Existing snake and insect robots can crawl in pipes and operate within centimeter-scale spaces; however, they are too slow or not completely autonomous. There is a pressing need for fast, agile, and efficient autonomous systems specifically designed for crawlspaces. Legged locomotion and multi-rotor flight show promise in addressing this challenge, having already revolutionized numerous remote sensing tasks by surpassing the capabilities of any other robot category. However, their operation remains limited to spacious tunnels. This Faculty Early Career Development (CAREER) project supports research that seeks to harness the exceptional agility, efficiency, and speed of legged and rotary-wing robots and adapt it specifically for use in crawlspaces. The project designs a multi-modal robot with extensive locomotion plasticity capable of efficiently traversing extremely tight crawlspaces with agility through bipedal walking and flying. Furthermore, this project aims to promote gender equity in Northeastern University’s robotics program, a significant area of growth identified by school leadership.Achieving efficient, agile, and fast legged-aerial locomotion in crawlspaces represents a new advancement in robot locomotion. Three knowledge gaps still exist in bridging legged locomotion and multi-rotor flight in crawlspaces: (1) Actuation challenges hinder the scalability of motion control performance, necessary for fast and precise foot placement, when transitioning from legged robots designed for open spaces to smaller robots operating in crawlspaces; (2) Instability issues arising from multi-rotors’ air jets near surfaces pose flight immobilization risks; (3) Crawlspaces need several modes of locomotion and there is no systematic robot design framework to accommodate the requirements dictated by these modes. This project will engage in fundamental research to address these gaps, designing a bioinspired locomotion robot capable of walking, hovering, jumping, and running over inclined surfaces to push the operational envelope of mobile robots, making autonomous operations inside extremely tight crawlspaces possible. The project’s efforts revolve around robot and control design. This project’s key areas of innovation include (1) Introducing actuation design paradigms for small robots based on computational structure design for achieving comparable motion control performance seen in large robots; (2) Research and validation of new locomotion feats and underlying models and nonlinear controllers based on the integration of posture manipulation and thrust vectoring to overcome air jet risks in aerial robotics; (3) Co-designing robots and controls through generative design methods to accommodate conflicting requirements imposed by many locomotion modes.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.