EFRI C3 SoRo: Overcoming Challenges in Control of Continuum Soft Robots through Data-driven Dynamic Decomposition and Light-modulated Materials

点击翻译按钮获取中文摘要

This project will advance the control of continuum soft robots in complementary areas of representation and implementation. First, it will explore the use of data-driven modeling techniques that are especially well-suited to high-dimensional nonlinear systems. Continuum systems are high dimensional because they may be formed into many distinct, independent shapes, while soft structures frequently exhibit nonlinear dynamics due to their characteristically large deformations. Second, it will explore sensing and actuation using materials that change their mechanical properties in response to light. Light may be steered to various parts of the robot through waveguides made from compliant materials in a variety of geometries, chosen to naturally match the compliance and geometry of the robot. Multiple commands or measurements may be combined on a single waveguide by using different colors of light for different signals. This helps address the large number of sensors and actuators needed to fully monitor and control these high-dimensional systems. The results of this work will be a new class of optically controllable, continuum, compliant, and configurable robots, or C4 optorobots for short. The capabilities of C4 optorobots will be valuable in numerous applications, including inspection of hard-to-reach areas and minimally invasive surgery. This project will include validation experiments, such as a representative engine inspection task, simulated tissue ablation in the heart, and simulated blood clot removal in the brain. Outreach activities of this project will focus on training of graduate and undergraduate students from underrepresented groups, including community college students, in emerging areas of soft robotics.This project will combine two key innovations - Koopman Operator Theory (KOT) and light-modulated materials design. KOT uses a data-driven approach to identify the eigenmodes of a complex, nonlinear system, thus enabling optimization of control inputs for greatest impact. KOT can produce a linear dynamical model that matches the system behavior, suitable for use in a model-based control. Light combined with photoresponsive materials can provide such control inputs because wavelength, intensity, pulse duration, and spatial distribution can be precisely and nearly instantaneously controlled. Accordingly, this project will accomplish the following goals: 1) Use KOT to extract and formulate linear, dynamic models of C3 robotic sub-systems and implement model-based control schemes using finite control inputs; 2) Develop transduction mechanisms, through engineered light delivery, novel chemical synthesis, and integration into light-responsive materials, to actuate C3 robotic sub-systems; 3) Integrate light-activated robotic sub-systems and model-based linear control schemes to realize controllable, continuum, compliant, and configurable robots -- C4 optorobots.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.