AI based design assistance system for soft robotics -Optimizing complex systems based on the smallest design entity

在自然界中，许多动物具有优于僵硬机器人结构的能力，无论或因为它们的顺从性。与传统的硬机器人相比，软机器人具有全新的和互补的能力，这在很大程度上形成了机器人学的一个新兴研究领域。然而，为了开发它们的适应性、灵活性和柔顺性，需要全新的设计和控制方法。本项目的目的是开发一个设计辅助系统，该系统基于对软机器人系统的最小设计实体-万能软弯曲执行器的加压硅胶气缸-的深刻知识，以确定在多参数空间内的几何、驱动和控制的最优设计。设计辅助系统将基于基本模型和正向运动学和动力学分析，并借助物理模型和机器学习算法的分层方法。通过定义万能弯曲致动器应满足的期望路径和作用在结构上的载荷，设计辅助系统应确定所有组件的最佳尺寸，以及串联连接的致动器的数量及其致动。附加信息，如弯曲执行器中的应力分布、静态和动态变形和方向，以及工作空间应作为输出。在第一个资助期，重点放在以通用软弯曲执行器为例的适当训练数据的数值生成、设计辅助系统的建立和不同机器学习算法的评估，而在第二个资助期，应进一步纳入一般设计原则、由于制造错误引起的不确定性、不特定的或与执行有关的参数以及对障碍物的控制。尽管底层的物理模型和机器学习方法适合于软机器人的应用，但基于系统中最小设计实体建立模块化设计辅助系统的一般方法可以转移到许多其他设计问题上。

In nature, many animals have capabilities that are superior to a rigid robotic structure, despite or because of their compliance. Largely because of their completely new and complementary capabilities compared to conventional hard robots, soft robots form an emerging field of research in robotics. However, to exploit their adaptability, flexibility and compliance, entirely new design and control approaches are needed.The aim of this project is to develop a design assistance system based on profound knowledge of the smallest design entity of a soft robotic system - the pressurizable silicone cylinders of a universal soft bending actuator - to define the optimal design within the multi-parameter space in geometry, actuation and control. The design assistance system will be based on underlying models and analysis of the forward kinematics and dynamics with help of a hierarchical approach of physical models and machine learning algorithms. By just defining the desired path that the universal bending actuator shall fulfill and the loads acting on the structure, the design assistance system shall determine the optimal dimension of all components, as well as the number of actuators connected in series and their actuation. Additional information, as the stress distribution in the bending actuator, the static and dynamic deformation and orientation, and the workspace shall be given as an output. In the first funding period, the focus lies on the numerical generation of appropriate training data for the example of the universal soft bending actuator, the establishment of the design assistance system and the evaluation of different machine learning algorithms, while in the second funding period further general design principals, uncertainties due to fabrication errors, unspecific or actuation dependant parameters, and the control around obstacles shall be incorporated. Even though the underlying physical models and machine learning approaches are adapted to the application of soft robotics, the general approach to build up a modular design assistance system based on the smallest design entity in the system can be transferred to many other design problems.