About Me

My name is Yuhao Jiang (蒋宇豪). I'm a postdoc researcher working with Prof. Jamie Paik at Reconfigurable Robotics Lab, EPFL. I obtained my Ph.D. degree in Mechanical Engineering from Arizona State University in Aug. 2023 under the supervision of Prof. Daniel Aukes. I received my M.S. and B.S. degree in Mechanical Engineering respectively from University of Florida in 2017 and Donghua University in 2015.

Research Interest

My research focuses on reconfigurable modular robotic systems that adapt their physical embodiment to changing tasks and environments. On the hardware side, I study mechanical intelligence through the co-design of mechanisms, materials, actuation, transmission, and embedded sensing. I develop compliant, multistable, and variable-stiffness mechanisms that generate complex motions and interactions with fewer actuators, addressing a central challenge in robotics: achieving versatile behavior without simply adding more actuators and increasingly complex controllers. On the software side, I study embodied intelligence through advanced multi-module embodiment planning, where the morphology and configuration of a reconfigurable modular robotic system are treated as planning variables that can be optimized in real time according to the task, environment, and physical constraints. Together, my work aims to create a physical AI ecosystem in which distributed modular robots gather rich real-world interaction data, while evolving AI models improve the robots’ coordination and task-level capabilities, laying a foundation for embodied artificial general intelligence in the physical world.

Selected Projects

CPG-Based Multi-Module Robotic Manipulation

This project marks the first application of the Central Pattern Generator (CPG) method to manipulation tasks in multi-module robotic systems. A novel data-driven optimization framework, integrated with dynamic simulations, efficiently identifies optimal CPG parameters in complex, high-dimensional spaces. This enables robust manipulation of objects with diverse sizes, shapes, and materials, as demonstrated through simulations and prototype experiments.

Soft Twisted Beam Vibration for Robotic Walking

This project investigates a novel, underactuated soft twisted beam structure that leverages structural intelligence to achieve effective locomotion. When actuated by a simple 2D vibrational force, the structure generates complex 3D motions. The study first examines the dynamic behavior of a single twisted beam, and then integrates these beams into a quadrupedal robot, Flix_Walker, which demonstrates three distinct and versatile locomotion modes, all achieved with minimal actuation.

Shape Change Propagation for Robotic Swimming

This project introduces a soft tubular swimming robot that leverages shape propagation—a concept where actuation forces are transformed by the robot's changing shape. By transmitting deformation from a central soft pneumatic actuator to curved fins, the device produces complex, asymmetric swimming strokes from simple, symmetric inputs. This approach simplifies control and power delivery while utilizing the mechanics of curved soft materials for effective locomotion.

Reconfigurable Curved Beams for Robotic Swimming

This project explores how the stiffness and buckling behavior of curved beams can be tuned through changes in length, camber angle, and width. By leveraging the natural tendency of curved beams to preferentially buckle in specific directions, the design enables passive generation of net work and moments from simple, symmetric inputs—reducing the need for complex control in soft robotic systems.

Bio-inspired Design for Robotic Swimming

This work introduces a fish-inspired underwater robot for maneuverability in open-channel canals under external disturbances. A machine learning workflow is used to train and select efficient swimming gaits in the lab, minimizing outdoor data collection by transferring only the most promising gaits for real-world testing. The key contribution is an online learning strategy that reliably identifies gaits with consistent performance across both laboratory and real-world environments.

Reconfigurable Soft Hinges via Pinched Tubes

This project introduces reconfigurable soft joints created by pinching thin-walled cylindrical tubes. These "virtual hinges" offer tunable compliance and can recover their original shape and stiffness when released. By leveraging 3D printing for rapid prototyping, the design enables easy adjustment of tube geometry and stiffness, allowing for customizable, passive rotational joints in soft robotic systems.

Public Outreach

Media Interview

1. RTS Education and Scientific Program: Feature in “A guide to the future: Swiss Federal Institute of Technology 02”
   Watch on YouTube (from 9:21)

Organized Events

1. IROS 2025 Workshop:
   S’MORE: Shape-Morphing Robotics via Embodied Sensing and Mechanisms Website
2. 2024 RRL Demo Day:
   Full-day public event for projects from RRL and ME-410 class
   Website, Report
3. 2023 RRL Demo Day:
   Full-day public event for projects from RRL and ME-410 class
   Website, Report
4. Robosoft 2021 Workshop:
   “Breaking the Mold: Challenging Current Paradigms in Soft Robotics”
   Website

Demos and Expositions

1. RRL lab tours (~6 times per year)
2. 2024 RRL Demo Day
3. 2024 Swiss Robotics Day
4. 2023 RRL Demo Day
5. 2023 Swiss Robotics Day
6. IdeaLab lab tours (~4 times per year)
7. 2019 Southwest Robotics Symposium (SWRS)

Academic Services

• Editorial Services: IROS 2026 Associate Editor.
• Journal Reviewer: The International Journal of Robotics Research (IJRR), IEEE Transactions on Robotics (T-RO), IEEE Robotics and Automation Letters (RA-L), Soft Robotics (SoRo), IEEE/ASME Transactions on Mechatronics (T-Mech), Journal of Field Robotics (JFR), ASME Journal of Mechanisms and Robotics (JMR).
• Conference Reviewer: IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), International Conference on Robotics and Automation (ICRA), International Conference on Soft Robotics (Robosoft), ACM Symposium on Computational Fabrication (SCF)