[Paper Review] Design Iterations for Passive Aerial Manipulator

B. V. Vidyadhara, Lima Agnel Tony+6 more|arXiv (Cornell University)|Feb 16, 2021

Robotic Path Planning Algorithms11 references4 citations

TL;DR

This paper presents a lightweight, energy-efficient passive aerial manipulator with a single-degree-of-freedom basket-type end-effector for intercepting moving targets, such as a 150 g ball up to 20 cm in diameter. The system features a novel in-flight extending propeller guard for drone safety and is validated through prototyping and field testing on a hexacopter platform with modular, low-power components.

ABSTRACT

Grabbing a manoeuvring target using drones is a challenging problem. This paper presents the design, development, and prototyping of a novel aerial manipulator for target interception. It is a single Degree of Freedom (DoF) manipulator with passive basket-type end-effector. The proposed design is energy efficient, light weight and suitable for aerial grabbing applications. The detailed design of the proposed manipulation mechanism and a novel in-flight extending propeller guard, is reported in this paper.

Motivation & Objective

• To design and prototype a lightweight, energy-efficient aerial manipulator for intercepting maneuvering targets in dynamic environments.
• To address safety challenges during target interception, particularly propeller damage from collisions.
• To develop a passive, low-power end-effector capable of reliably grasping spherical objects up to 20 cm in diameter.
• To integrate a novel in-flight extending propeller guard mechanism that deploys during manipulation to protect the drone.
• To ensure the entire system remains within a 120×120×50 cm volume during takeoff and landing for operational compactness.

Proposed method

• The end-effector uses a passive basket-type mechanism with interlocking prongs to capture spherical targets without actuation, reducing power consumption.
• A series of design iterations were conducted, including active interlocking claws and passive basket prototypes, to optimize for weight, strength, and grab reliability.
• The manipulator arm is made of carbon fiber and aluminum, with a 1.1 m extension, mounted on a hexacopter (M600) with three counter-motors to balance moments.
• An extendable propeller guard was designed using 3D-printed standoffs and carbon fiber guide rods, actuated via a continuous rotation servo to deploy during flight.
• The system integrates a DJI A3 Pro autopilot with triple-redundant GPS and an Nvidia Jetson TX2 for on-board computation, with Arduino Mega for actuation control.
• Field testing was conducted at IISc airfield to validate performance, with adjustments made based on observed failures and inefficiencies.

Experimental results

Research questions

• RQ1 How can a passive, low-power aerial manipulator be designed to reliably intercept and capture a moving target up to 150 g and 20 cm in diameter?
• RQ2 What mechanical design enables safe in-flight deployment of a propeller guard without compromising drone stability or maneuverability?
• RQ3 How do structural and geometric choices in the end-effector affect grab success rate and footprint?
• RQ4 What are the trade-offs between material selection, weight, and rigidity in a lightweight aerial manipulator arm?
• RQ5 How can a modular, energy-efficient, and compact aerial grasping system be developed for diverse applications like fruit picking or counter-drone missions?

Key findings

• The final passive basket-type end-effector successfully captured targets up to 150 g and 20 cm in diameter, with minimal energy consumption.
• The in-flight extending propeller guard provided sufficient clearance and protected the propellers during interception attempts.
• Field tests confirmed the system's robustness under mild wind disturbances and demonstrated reliable target interception.
• The use of carbon fiber and birch wood in the end-effector provided sufficient strength for repeated detachment cycles.
• Design iterations revealed that prong geometry and attachment point placement significantly influence grab success, suggesting room for optimization.
• Piezoelectric sensors were identified as a more energy-efficient alternative to the current limit-switch-based grab detection system.