CAREER: Adaptive Actuation and Control in Embodied Biohybrid Robots

动物通常是机器人设计的灵感来源。通过根据动物的蓝图进行设计，工程师们可以创造出能够行走、奔跑、爬行、游泳甚至飞行的机器人系统。然而，即使机器人技术在过去的几十年里取得了进步，机器人系统仍然没有在生物动物身上看到的许多能力。现有机器人与动物机器人之间的一个关键区别是，生物系统是由柔软的、适应性强的材料组成的，包括用于驱动的肌肉和用于控制的神经元。该职业奖研究如何使用活肌肉为生物混合机器人制造鲁棒，适应性强的执行器，这些执行器如何适应运动，以及如何控制具有活神经元的生物混合机器人。此外，该职业奖支持教育和推广计划，以改善机器人和STEM领域不同学生和教师的招聘和保留。根据研究成果，将开发适合年龄的教育材料，提供给初中和高中教师，并整合到生物启发机器人的研究生课程中。研究团队将举办虚拟和面对面的推广活动，向学生介绍生物启发和生物混合机器人。未被充分代表的本科生将被招募参加生物混合机器人和建模的暑期研究经验。最后，研究者将推广机器人领域女性教师的招聘和保留工具。这个为期5年的CAREER项目将使生物致动器能够通过类似肌腱的界面与一系列机器人结构连接，生物致动器控制的生物启发神经网络，以及执行生物混合机器人基本“编程”的能力。生物混合机器人直接利用活体组织作为可再生工程材料。特别是，基于肌肉的生物致动器具有自我修复、柔顺和适应负载的能力。到目前为止，大多数生物混合研究都集中在作为系统单个组成部分的生物材料上，需要集成设计、制造和“编程”鲁棒生物执行器和生物控制网络的方法来提高生物混合机器人的性能和扩大适用性。为了满足这一需求，这个CAREER项目将(1)通过创建嵌入式生物兼容接口，实现广泛的机器人外围设备的自适应生物驱动，(2)建模和制造简单的生物神经网络来控制生物执行器，(3)训练集成的生物执行器和生物神经网络。所提出的研究方法不仅将导致生物驱动和控制的进步，而且还将特别关注集成生物混合机器人的开发。“可编程”生物混合机器人在医学上有应用，小型生物相容性系统可以用作自驱动支架或医疗植入物，或者作为药物筛选和神经科学芯片上神经肌肉组织的功能组件。提出的研究为解决生物混合机器人的未来挑战奠定了基础，包括将多种传感模式集成到生物混合机器人系统中，理解体现对神经肌肉控制回路的影响，以及研究分布式生物混合驱动系统中的紧急动力学。本CAREER提案中的研究方法将与教育和推广计划相结合，以(1)将神经肌肉建模纳入生物混合机器人课程，(2)通过生物混合机器人体验提高不同学生在机器人领域的保留率，以及(3)建立工具来提高女性教师在机器人领域的知名度，以提高保留率。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。

Animals are often a source of inspiration in robotic design. By designing from animal blueprints, engineers can create robotic systems capable of walking, running, crawling, swimming, and even flying. However, even with the advances in robotics over the past decades, robotic systems still fall short of many of the capabilities seen in biological animals. One key difference between existing robots and their animal counterparts is that biological systems are made up of soft, adaptable materials, including muscles for actuation and neurons for control. This CAREER award investigates how to fabricate robust, adaptable actuators for biohybrid robots using living muscle, how these actuators adapt to exercise, and how to control biohybrid robots with living neurons. Additionally, this CAREER award supports educational and outreach initiatives to improve recruitment and retention of diverse students and faculty in robotics and STEM. Accessible age-appropriate educational materials based on the research outcomes will be developed, made available to middle and high school teachers, and integrated into a graduate course on Bioinspired Robotics. The research team will host virtual and in-person outreach events to introduce students to bioinspired and biohybrid robotics. Underrepresented undergraduate students will be recruited for summer research experiences in biohybrid robotics and modeling. Finally, the investigator will promote tools for recruitment and retention of women faculty in robotics. This 5-year CAREER project will result in bioactuators capable of interfacing with a range of robotic structures via tendon-like interfaces, bioinspired neural networks for bioactuator control, and the ability to perform basic ‘programming’ of biohybrid robots. Biohybrid robotics directly harnesses living tissues as renewable engineering materials. In particular, muscle-based bioactuators are self-healing, compliant, and adapt to loading. Whereas most biohybrid research to date has focused on biological materials as individual components of the system, approaches for the integrated design, fabrication, and ‘programming’ of robust bioactuators and biological control networks are needed to improve biohybrid robot performance and broaden applicability. To meet this need, this CAREER project will (1) enable adaptive bioactuation of a wide range of robotic peripheries through the creation of embedded biocompatible interfaces, (2) model and fabricate simple biological neural networks to control bioactuators, and (3) train integrated bioactuators and biological neural networks. Not only will the proposed research approach lead to advances in bioactuation and control, but it will also specifically focus on integrated biohybrid robot development. ‘Programmable’ biohybrid robots have applications in medicine where small-scale biocompatible systems could be used as self-actuating stents or medical implants, or as functional components of neuromuscular tissues-on-a-chip for drug-screening and neuroscience. The proposed research lays the foundation for addressing future challenges in biohybrid robotics, including integrating diverse sensing modalities into biohybrid robot systems, understanding the effect of embodiment on neuromuscular control circuits, and studying emergent dynamics in distributed biohybrid actuation systems. The research approach in this CAREER proposal will be integrated with an educational and outreach plan to (1) incorporate neuromuscular modeling in biohybrid robotics curriculum, (2) improve retention of diverse students in robotics through biohybrid robot experiences, and (3) build tools to improve visibility of women faculty in robotics towards improving retention.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.