Morpho-functional Machines with Manifold Actuated Joints
Reinforcement Learning-Based Model Matching in COBRA, a Slithering Snake Robot
This research addresses the “sim-to-real problem” encountered in many robotic systems, but more so in those with contact-rich dynamics such as our snake robot COBRA. In this research, a reinforcement learning-based model identification approach is employed aimed at enhancing the accuracy of contact dynamics in simulation for COBRA. Through iterative optimization, RL model refines the parameters of COBRA’s dynamical model such as coefficient of friction and actuator parameters using experimental and simulated data.
Snake Robot with Tactile Perception Navigates on Large-scale Challenging Terrain
With advancements in robot skin technology, snake robots now feature body-surface tactile perception. We proposed a locomotion control framework that leverages this perception to enhance adaptability across various terrains. Using a hierarchical reinforcement learning (HRL) architecture, high-level controls manage global navigation, while low-level controls employ curriculum learning for local maneuvers. The computational demands of collision detection in whole-body tactile sensing reduce simulator efficiency. To address this, we implemented a distributed training pattern. Our evaluation in large-scale cave exploration demonstrated improved motion efficiency, confirming the benefits of tactile perception in terrain-adaptive locomotion.
Dynamic Posture Manipulation During Tumbling for Closed-Loop Heading Angle Control
Passive tumbling uses natural forces like gravity for efficient travel. But without an active means of control, passive tumblers must rely entirely on external forces. Northeastern University’s COBRA is a snake robot that can morph into a ring, which employs passive tumbling to traverse down slopes. However, due to its articulated joints, it is also capable of dynamically altering its posture to manipulate the dynamics of the tumbling locomotion for active steering. This paper presents a modelling and control strategy based on collocation optimization for real-time steering of COBRA’s tumbling locomotion. We validate our approach using Matlab simulation
Non-impulsive Contact-Implicit Motion Planning for Morpho-functional Loco-manipulation
Object manipulation has been extensively studied in the context of fixed base and mobile manipulators. However, the overactuated locomotion modality employed by snake robots allows for a unique blend of object manipulation through locomotion, referred to as loco-manipulation. The following work presents an optimization approach to solving the loco- manipulation problem based on non-impulsive implicit contact path planning for our snake robot COBRA. We present the mathematical framework and show high fidelity simulation results for fixed-shape lateral rolling trajectories that demonstrate the object manipulation.
Loco-Manipulation with Nonimpulsive Contact-Implicit Planning in a Slithering Robot
Object manipulation has been extensively studied in the context of fixed base and mobile manipulators. However, the overactuated locomotion modality employed by snake robots allows for a unique blend of object manipulation through locomotion, referred to as loco-manipulation. The following work presents an optimization approach to solving the loco- manipulation problem based on non-impulsive implicit contact path planning for our snake robot COBRA. We present the mathematical framework and show high-fidelity simulation results and experiments to demonstrate the effectiveness of our approach.
Hierarchical RL-Guided Large-scale Navigation of a Snake Robot
Classical snake robot control leverages mimicking snake-like gaits tuned for specific environments. However, to operate adaptively in unstructured environments, gait generation must be dynamically scheduled. In this work, we present a four- layer hierarchical control scheme to enable the snake robot to navigate freely in large-scale environments. The proposed model decomposes navigation into global planning, local planning, gait generation, and gait tracking. Using reinforcement learning (RL) and a central pattern generator (CPG), our method learns to navigate in complex mazes within hours and can be directly deployed to arbitrary new environments in a zero-shot fashion. We use the high-fidelity model of Northeastern’s slithering robot COBRA to test the effectiveness of the proposed hierarchical control approach.
How Strong a Kick Should be to Topple Northeastern’s Tumbling Robot?
Rough terrain locomotion has remained one of the most challenging mobility questions. In 2022, NASA’s Innovative Advanced Concepts (NIAC) Program invited US academic institutions to participate NASA’s Breakthrough, Innovative &Game-changing (BIG) Idea competition by proposing novel mobility systems that can negotiate extremely rough terrain, lunar bumpy craters. In this competition, Northeastern University won NASA’s top Artemis Award award by proposing an articulated robot tumbler called COBRA (Crater Observing Bio-inspired Rolling Articulator). This research studies the inherent stability of the tumbling structure due to gyroscopic effects and models the dynamics of tumbling locomotion towards closed loop control.
Heading Control for Obstacle Avoidance using Dynamic Posture Manipulation during Tumbling Locomotion
Passive tumbling structures are energy efficient, but often sacrifice control authority due to their under-actuated nature. Unlike many passive tumbling robots, Northeastern University’s COBRA is a snake robot with eleven articulated joints that transforms into a wheel-like structure with a high degree of posture control during tumbling, and using this posture manipulation, COBRA can control its forward velocity and heading angle while tumbling. This research develops the mathematical framework that describes the dynamics of posture manipulation during tumbling and identifies two types of control actions that allow it to control its movement. This goal is to achieve desired trajectory tracking during passive tumbling using only posture manipulation.