NRI: FND: Robust and Scalable Planning for Agile and Collaborative Robot Teammates in Complex Environments

有一天，协作机器人团队可能会在我们的生活中无处不在，在充满挑战的环境中陪伴人类，执行各种任务，甚至在需要时超越人类的能力。这种潜力为协作机器人团队在日常生活和救援任务中与人类有效可靠地合作提供了重要机会。然而，设计机器人决策和规划算法仍然具有挑战性，这些算法可扩展到复杂的现实世界场景，并提供正式的安全保证。正式的决策方法对于推断机器人队友如何在非结构化环境中协作至关重要，例如，地面机器人如何与空中机器人协调搜索和救援任务，或者人形机器人如何与人类合作在密闭空间中提供物资。当机器人被要求达到或超过人类水平的灵活性和熟练度时，这一挑战将更加突出。这个跨学科项目的总体愿景是实现在复杂环境中无处不在的协作机器人机动的鲁棒和可扩展算法，同时有效地与人类合作完成各种任务。广泛的应用将被瞄准，包括家庭护理应用和搜索和救援任务。虽然该项目侧重于腿式机器人和空中机器人作为设计示例，但该方法的目标是更广泛的机器人系统，包括轮式机器人、操纵器和现场环境，如网络控制系统和运输系统。该项目还包括一项综合教育计划，包括开设一门关于机器人系统正式控制方法的新研究生课程，以及针对女性和代表性不足的少数民族学生的基于stem的外展倡议。该项目旨在推进异构和协作机器人系统的规划和决策算法，以在非结构化、人在场景和动态变化的环境中实现无处不在的任务。该项目将针对以下三个目标：设计基于动态规划和控制障碍证书的鲁棒非周期性运动规划器，用于多用途地面和空中机动；综合博弈论，反应性和鲁棒性任务计划，以应对不同的环境事件；提出了一种新的多智能体决策方法，将整个机器人团队分解为多个子团队。这个综合规划框架将采用最先进的形式化方法、多智能体系统、鲁棒控制和机器学习相互作用的算法方法。所提出的规划理论将侧重于鲁棒性和可扩展性推理，这是非常重要的，因为它们为实现复杂的多机器人可操作性和合作任务提供了机会，同时推理了形式保证，包括可证明的正确性和可保证的安全性。在计算资源充足的前提下，该框架能够在任务和任务规划两个层面实时地进行任务决策并生成满足要求的运动计划。该项目的可交付成果包括基于开源算法和实验实现的统一、异构腿式和空中机器人的决策和规划理论。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。

Collaborative robot teams may one day become ubiquitous in our lives, accompany humans in challenging environments, perform a wide range of task, and even surpsass human capabilities where needed. This potential offers a significant opportunity for co-robot teams to collaborate effectively and reliably with humans for daily life and rescue tasks. However, it remains challenging to design robotic decision-making and planning algorithms that are scalable to complex real world scenarios and that provide formal safety assurances. Formal decision-making methods are critical in reasoning about how robot teammates collaborate in unstructured environments, e.g., how a ground robot coordinates with aerial robots for search and rescue tasks, or how a humanoid robot cooperates with humans to deliver supplies in confined space. This challenge will be accentuated when robots are required to perform tasks at or beyond human-level dexterity and proficiency. The overarching vision of this interdisciplinary project is to achieve robust and scalable algorithms for ubiquitous co-robots maneuvering in complex environments while effectively collaborating with humans for various tasks. A wide range of applications will be targeted, including home care applications and search and rescue tasks. Although this project focuses on legged and aerial robots as design examples, the approaches are targeting a broader range of robotic systems including wheeled robots, manipulators, and fielding environments such as networked control systems and transportation systems. This project also features an integrated education plan involving the creation of a new graduate course on formal control methods for robotic systems as well as STEM-based outreach initiative for women and underrepresented minority students.This project aims at advancing planning and decision-making algorithms of heterogeneous and collaborative robotic systems to achieve ubiquitous tasks in unstructured, human-in-the-scene, and dynamically changing environments. This project will target the following three objectives: devise robust, non-periodic motion planners based on kinodynamic planning and control barrier certificates for versatile terrestrial and aerial maneuvering; synthesize game-theoretic, reactive, and robust task planners in response to diverse environmental events; and propose a novel multi-agent decision-making approach decomposing the whole robot team into multiple sub-teams. This integrated planning framework will adopt algorithmic methods at the interaction of state-of-the-art formal methods, multi-agent systems, robust control, and machine learning. The proposed planning theory will focus on robustness and scalability reasoning which are of high importance, as they open up the opportunity for achieving complex multi-robot maneuverability and cooperation tasks while reasoning about formal guarantees including provable correctness and assured safety. Under the hypothesis of sufficient computational resources, the proposed framework will make task decisions and generate motion plans in real-time satisfying the required specifications at both mission and task planning levels. The deliverables from this project include decision-making and planning theories envisioned for unified, heterogeneous legged and aerial robots with open-sourced algorithms and experiment implementations.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.