Source-linked AI summary
Vision-Guided Morphing Quadcopter for Multi-Geometry Payload Transport through Narrow Passages
Aashish Sahu, Shriram Hari, R. Prasanth Kumar
TL;DR
Aerial payload transport is challenging when payload geometry, grasp stability, flight control, and passage width vary together. The paper presents a vision-guided morphing quadcopter using a single-actuator caging mechanism, and simulation completes transport for three payload geometries while substantially reducing footprint.
Problem
Aerial transport requires grasping payloads with different geometries while maintaining flight stability and traversing narrow passages.
Method
A quadcopter uses four hybrid arm–leg structures, onboard vision, endpoint force feedback, phase-wise planning, and a single tendon-driven actuator for adaptive grasping and footprint morphing.
Results
0.31 m maximum RMS position error and 75.0–89.7% footprint reduction were achieved across box, cylindrical, and spherical payload simulations.
Takeaways & Limitations
The single-actuator morphology-adaptive design provided shape-tolerant payload confinement while reducing the effective vehicle footprint for constrained-space transport.
Abstract
from arXiv · showhide
Aerial payload transport using multirotor unmanned aerial vehicles is challenging because payload geometry, contact interaction, grasp stability, flight control, and narrow-passage traversal are strongly coupled during pickup and transport. Object-specific grippers often cannot adapt their footprint or grasp geometry when the payload shape or passage width changes. This paper presents a vision-guided morphing quadcopter for multi-geometry payload transport through narrow passages. The proposed platform uses four hybrid arm-leg structures that function as both landing supports and grasping members. A centrally placed actuator drives a tendon-based morphing mechanism, enabling all four arms to synchronously retract or expand for object grasping, footprint reduction, and post-transport release. Onboard vision estimates the payload geometry and passage width, while endpoint force feedback is used to confirm grasp contact during payload engagement. A phase-wise mission planner, PID-based flight stabilization, and morphology-adaptive grasp controller are implemented in a MuJoCo simulation environment. The framework is evaluated using box, cylindrical, and spherical payloads, representing flat-faced, rolling-curved, and fully curved contact conditions. Across the three cases, the simulated system completes the pickup-transport-release sequence with a maximum RMS position error of 0.31 m, a final drop-zone error below 0.18 m, a compact grasp footprint of 0.09-0.21 m2, and a footprint reduction of 75.0-89.7 percent. The results demonstrate that a single-actuator morphing quadcopter can adapt its grasp footprint for the transport of payloads with different geometries while reducing its overall footprint for narrow-passage traversal.
I. INTRODUCTION
Aerial payload transport is difficult because flight stability, contact, grasping, payload geometry, and narrow-passage traversal are coupled. The paper addresses these constraints with a vision-guided quadcopter whose single morphing mechanism supports multi-geometry grasping and footprint reduction.
- Aerial grasping couples stable flight with payload contact, grasp force, center-of-mass variation, rotor disturbances, and navigation constraints during transport.
- Conventional aerial graspers often depend on known grasp points, specific payload geometries, or dedicated mechanisms, limiting reuse across object shapes.
- Fixed quadcopter footprints can restrict safe traversal through windows, corridors, door frames, collapsed structures, and industrial openings.
- Geometric caging constrains objects within a controlled contact envelope, offering a grasping principle suited to contact uncertainty and rotor disturbance.
- A centrally actuated tendon mechanism synchronously retracts or expands four hybrid arm–leg structures for grasping, footprint reduction, and release.
- The integrated framework combines onboard vision, endpoint force feedback, phase-wise trajectory generation, morphology-adaptive grasping, and PID-based flight stabilization in MuJoCo.
II. SYSTEM ARCHITECTURE
The system architecture combines hybrid arm–leg structures, onboard perception, mission sequencing, morphology adaptation, and flight control. A single central actuator changes the quadcopter between open and compact configurations for grasping and narrow-passage transport.
- SYSTEM ARCHITECTURE: Four hybrid arm–leg structures serve as landing supports during touchdown and grasping members during payload engagement.
- SYSTEM ARCHITECTURE: A centrally located motor synchronously retracts or expands all four members, changing the vehicle’s effective footprint with one actuator.
- SYSTEM ARCHITECTURE: The MuJoCo framework models the quadcopter, morphing actuator, arm–leg members, contact interaction, corridor geometry, and box, cylinder, and sphere payloads.
- SYSTEM ARCHITECTURE: A downward-facing camera estimates payload size and shape, while a front-facing camera estimates passage width for configuration selection.
- SYSTEM ARCHITECTURE: Endpoint force sensors confirm contact and sufficient payload confinement during the compact caging-style grasp.
- SYSTEM ARCHITECTURE: θm(t) = θmaxs(t) maps the mission-phase transition variable to the commanded morphing angle; decreasing s(t) during release expands the structure.
- SYSTEM ARCHITECTURE: Reduced effective arm radius and grasp-footprint area indicate compact morphology supporting payload confinement and narrow-passage footprint reduction.
III. METHODOLOGY AND CONTROL DESIGN
The methodology combines vision-based localization, phase-wise trajectory generation, force-verified caging, morphology adaptation, and PID flight stabilization for payload transport through narrow passages.
- Mission planning: The mission planner generates sequential commands for takeoff, approach, alignment, descent, grasping, transport, release, and exit.
- Mission planning: A scouting circle followed by centering and descent reduces abrupt motion near the payload and improves alignment before contact.The reference trajectory uses the estimated payload location, scouting radius, and angular rate.
- Perception and morphology: Camera measurements of payload dimensions and passage width determine whether the quadcopter remains open or reduces its effective footprint.The downward-facing camera estimates payload size and shape, while the front-facing camera estimates the gap between obstacle boundaries.
- Perception and morphology: A single central motor synchronously retracts or expands four hybrid arm–leg members, forming a compact caging grasp during transport and reopening during release.The mechanism supports open flight, compact grasping, narrow-passage traversal, and post-transport release.
- Grasp verification: Endpoint force feedback validates confinement within a safe force range, with positive grasp-force margin indicating sufficient contact for transport.The force condition is used with vision-based alignment as logical grasp confirmation rather than as a separate optimization objective.
- Evaluation: The evaluation measures tracking, drop-zone error, morphology reduction, passage clearance, and successful completion of pickup, transport, release, and exit.Success requires localization, compact grasp formation, narrow-passage transport, goal-region release, and stable exit.
IV. SIMULATION RESULTS AND DISCUSSION
MuJoCo simulations show that the morphing quadcopter completes payload transport across box, cylindrical, and spherical cases while adapting its morphology and grasp footprint for narrow-passage traversal.
- Simulation setup: The same mission sequence was executed for box, cylindrical, and spherical payloads, covering takeoff, alignment, grasping, constrained transport, release, and exit.The payloads represent flat-faced, rolling-curved, and fully curved contact conditions.
- Morphology adaptation: A compact caging-style grasp forms when the central motor synchronously retracts the four hybrid arm–leg members during grasping.The open configuration is retained during approach for rotor clearance and flight stability.
- Flight performance: The vehicle remains stable and completes the pickup–transport–release sequence for all payload geometries despite a small transient overshoot during grasp transport.The overshoot occurs while payload contact, morphology change, and inertial coupling happen simultaneously.
- Quantitative performance: 0.31 m maximum RMS position error and below 0.18 m final drop-zone error were achieved across the evaluated payloads.These values summarize the quantitative task-performance metrics.
- Footprint modulation: 75.0–89.7% footprint reduction was obtained as grasp-footprint area decreased from 0.84–0.87 m2 open to 0.09–0.21 m2 compact.The reduction reflects active changes in grasp geometry and effective vehicle footprint.
- Grasp confirmation: Endpoint force feedback verifies contact establishment before transport, preventing premature motion before payload engagement.The force condition serves as the grasp-confirmation criterion.
- Payload geometry: The box, cylinder, and sphere require different contact adaptations, yet the same hybrid arm–leg and single-actuator strategy transports all three.The box offers face and edge contact, the cylinder requires radial closure, and the sphere requires a wider compact grasp envelope.
V. CONCLUSION AND FUTURE WORK
The paper presents a single-actuator morphing quadcopter for multi-geometry payload transport through narrow passages. Simulations across three payload geometries show successful transport with reduced vehicle footprint, while future work targets physical validation and improved perception and control.
- Conclusion: Four hybrid arm–leg structures serve as both landing supports and grasping members, while one central actuator synchronously retracts or expands them.The tendon/thread mechanism supports grasping, footprint reduction, narrow-passage transport, and release.
- Conclusion: 0.31 m maximum RMS position error, below 0.18 m final drop-zone error, and 75.0–89.7% footprint reduction were achieved in MuJoCo simulations.The framework was evaluated with box, cylindrical, and spherical payloads.
- Future work: Future work includes depth-based perception, contact-aware control, improved force regulation, disturbance modeling, and physical prototype validation with varied payload properties.The stated validation scope includes different shapes, masses, and surface properties.