Research into resilient under-actuated robots

自2006年以来，工程系统的弹性得到了广泛的关注。韧性是一个系统通过自身的资源和力量从任何不幸中恢复的能力。提出这项研究的动机来自于我们对弹性系统，特别是弹性机器人系统重要性的理解。这项研究计划的长期目标是在系统，特别是机器人系统中建立弹性工程。为了确保研究成果得到最广泛的应用，拟议的研究将基于欠驱动机器人，这是最普遍和最复杂的机器人类型。这项研究的短期目标有四个：(1)设计具有弹性的欠驱动机器人，以实现尽可能多的恢复原则；(2)找到机器人的最佳配置，以恢复由于机器人部分损伤(如一条腿丢失)而失去的功能；(3)设计一个连接器，使主动、被动和固定的关节和脱节能够促进部分受损机器人的自我改变；(4)规划和调度自我改变过程，以实现机器人的最佳配置。 这项拟议的研究有望为具有弹性行为的欠驱动机器人系统的设计和操作提供理论和方法。该计划还将生产几个弹性机器人系统的原型。这项拟议的研究有几个智力上的优点。弹性欠驱动机器人的概念在文献中还没有被研究过，尽管它在通用性、能力和成本效益方面具有优势。事实上，人类有许多被动的关节和结构，动作不足。弹性机器人系统只具有某些结构的主动模块，是欠驱动机器人的一种特例，因此文献中针对欠驱动机器人所产生的知识适用于可自重构的自主模块化机器人。 机器人的弹性行为为社会增加了价值。例如，在战场上具有弹性的机器人可以有更高的机会战胜敌人，因为战场上作战能力交付的寿命肯定是一个关键因素。救援机器人在核电站事故和地震等灾难中的弹性行为是非常受欢迎的，因为在这些情况下，机器人需要交互的环境是动态的，机器人被损坏的机会更高。

Resilience of engineering systems has gained much attention since 2006. Resilience characterizes a system in terms of its ability to recover from any mishap through its own resources and power. The motivation of the proposed research comes from our understanding of the importance of resilient systems, especially resilient robotic systems. The long-term goal of this research program is to establish resilience engineering in systems, especially robotic systems. To ensure the research results will be most widely applicable, the proposed research will be based on under-actuated robots, which are the most general and the most complex type of robots. There are four short-term goals of the proposed research (in brief): (1) designing resilient under-actuated robots for enabling as many recovery principles as possible; (2) finding the best configuration of a robot to recover lost function due to partial damage of the robot (e.g. one leg lost); (3) designing a connector for enabling active, passive, and fixed joints and disjoints to facilitate self-changes of the partially damaged robot; (4) planning and scheduling the self-change process to achieve the best configuration of the robot. The proposed research is anticipated to produce theories and methodologies for the design and operation of under-actuated robotics systems with resilient behavior. The program will also produce several prototypes of resilient robotic systems. The proposed research has a couple of intellectual merits. The notion of the resilient under-actuated robot has not been studied in the literature despite its benefits in terms of generality, capability, and cost-effectiveness. In fact, humans are under-actuated with many passive joints and structures. Resilient robotic systems which only have active modules with some structures are a special case of the under-actuated robot, so the knowledge generated for under-actuated robots is applicable to self-reconfigurable autonomous modular robots in literature. The resilient behavior of robots adds value to society. For instance, resilient robots in a battlefield situation can have a higher chance to win against the enemy, as the longevity of combat capability delivery in battlefields is certainly a critical factor. The resilient behavior of rescuer robots for disasters such as nuclear power plant accidents and earthquakes is highly desired, because in those situations, the environment in which the robots need to interact is dynamic, and there is higher chance for the robots to get damaged.