TL;DR
This paper introduces SMORS, the first fully actuated soft multirotor UAV that combines rigid and soft arms for multimodal locomotion. By integrating soft manipulators directly into the aerial platform, SMORS achieves robust interactions and demonstrates energy-efficient jumping maneuvers, bridging the gap between soft and aerial robotics.
What Problem Does This Paper Address?
Traditional aerial robots, like hexarotors, are typically rigid. While they excel at aerial tasks, their rigidity can be a limitation when interacting with the environment, especially for manipulation or when encountering unexpected disturbances. Soft robots, on the other hand, offer inherent compliance and robustness, but integrating them with aerial platforms for complex locomotion has been challenging. Existing approaches often treat soft manipulators as separate payloads, leading to decoupled control and modeling. This paper addresses the need for an integrated aerial platform that leverages the benefits of soft robotics for enhanced interaction and multimodal locomotion, specifically aiming for robust physical contact and energy-efficient maneuvers like jumping.
Methods
The core innovation of SMORS lies in its hybrid design: a fully actuated hexarotor platform with three rigid arms and three continuously soft arms, each equipped with a propeller. The soft arms are actuated using a tendon-driven mechanism, allowing for controlled bending and variable tilting of the propellers. This design enables the platform to achieve full actuation, meaning it can independently control its position, orientation, and the forces/torques it exerts.
The authors developed a detailed dynamical model for SMORS, considering both the rigid main body and the soft arms. They employed a Piecewise Constant Curvature (PCC) model for the soft arms, which simplifies their kinematic and dynamic representation while maintaining accuracy. For control, they designed a closed-loop controller capable of tracking desired 3D poses and end-effector trajectories. This controller accounts for the complex dynamics of the soft arms and the interaction forces during contact.
Datasets
Not discussed in the paper. The experiments were conducted in simulation using MATLAB-Simulink with the Simscape toolbox.
Key Insights and Ideas
- Bridging Soft and Aerial Robotics: SMORS represents a significant step in integrating soft robotics directly into the core structure of an aerial platform, rather than treating soft components as mere add-ons. This integration allows for a more unified control and dynamic modeling.
- Multimodal Locomotion: The hybrid design enables novel locomotion capabilities, such as combining flight with jumping. The soft arms can store kinetic energy during ground contact, which is then released to aid in the jumping maneuver.
- Robust Interaction: The inherent compliance of the soft arms, combined with the platform's full actuation, leads to more robust interactions with the environment. This is crucial for tasks involving physical contact, where rigid robots might be prone to instability.
- Energy Efficiency: Simulations suggest that the multimodal locomotion approach, particularly the jumping maneuver, can be more energy-efficient than conventional flight, especially when assisted by gravity.
Key Results
- The paper presents the dynamical model and a closed-loop controller for the SMORS platform.
- Simulations demonstrate the platform's ability to track a jumping-flying trajectory with a peak norm position error of approximately 0.09m during contact and a peak norm attitude error of around 8°.
- The controller effectively manages the complex dynamics, including attitude changes during ground contact.
- Simulations indicate that the multimodal locomotion strategy can be more energy-efficient than pure hovering, with energy consumption decreasing as gravitational acceleration increases.
Limitations and Future Work
- Simulation-Based: All presented results are from simulations. A physical prototype is needed to validate the design and control approach in real-world scenarios.
- Simplified Soft Arm Model: While the PCC model is effective, it's a simplification of the true soft arm dynamics.
- Future Work: The authors plan to build a physical prototype, test the design and control, and develop optimal trajectory generation schemes that consider input saturation and minimize energy consumption for multimodal locomotion. They also aim to explore more complex interaction tasks and further investigate the energy efficiency benefits.
Summary updated