[Paper Review] Experimental Energy Consumption Analysis of a Flapping-Wing Robot

Raúl Tapia, Alvaro C. Satue+3 more|arXiv (Cornell University)|Jun 1, 2023

Biomimetic flight and propulsion mechanisms4 citations

TL;DR

This study experimentally analyzes energy consumption in a flapping-wing robot (E-Flap) across flight phases and onboard perception systems. It measures power use in launching, ascending, and cruising flight, and evaluates stereo and event-based cameras, finding event cameras consume less than 200 mW—significantly lower than conventional cameras—making them ideal for extending flight endurance in energy-constrained bioinspired drones.

ABSTRACT

One of the motivations for exploring flapping-wing aerial robotic systems is to seek energy reduction, by maintaining manoeuvrability, compared to conventional unmanned aerial systems. A Flapping Wing Flying Robot (FWFR) can glide in favourable wind conditions, decreasing energy consumption significantly. In addition, it is also necessary to investigate the power consumption of the components in the flapping-wing robot. In this work, two sets of the FWFR components are analyzed in terms of power consumption: a) motor/electronics components and b) a vision system for monitoring the environment during the flight. A measurement device is used to record the power utilization of the motors in the launching and ascending phases of the flight and also in cruising flight around the desired height. Additionally, an analysis of event cameras and stereo vision systems in terms of energy consumption has been performed. The results provide a first step towards decreasing battery usage and, consequently, providing additional flight time.

Motivation & Objective

• To quantify energy consumption across key flight phases—launching, ascending, and cruising—of a flapping-wing aerial robot (E-Flap).
• To evaluate the power consumption of onboard perception systems, particularly stereo and event-based cameras, in real-world flight-like conditions.
• To identify energy-efficient perception solutions that minimize power draw without compromising navigation or obstacle avoidance.
• To support longer flight endurance by optimizing component-level power usage in bioinspired aerial robots.
• To provide empirical data for predicting energy use in autonomous flapping-wing robots under different flight and sensor configurations.

Proposed method

• Conducted controlled flight tests in an indoor test bed to measure power draw during launching, ascending, and cruising phases using a custom measurement device.
• Used an INA219 wattmeter to record real-time electrical consumption of motors, servos, and electronics across flight phases.
• Mounted cameras (RealSense D435, ZED, eCapture G53, DAVIS346 DVS) on a pitch-oscillating benchmark to simulate flapping-wing motion.
• Measured power consumption under varying pitch rates (0–60°) and scene dynamics (rapid, slow, or static motion) to assess event rate dependency.
• Employed a VectorNav VN-200 sensor to track pitch rate and correlate it with camera power draw.
• Analyzed event generation rates (in Meps) for the DAVIS346 DVS and correlated them with power consumption to assess efficiency under dynamic scenes.

Experimental results

Research questions

• RQ1 What proportion of total energy is consumed during launching, ascending, and cruising phases in a flapping-wing robot?
• RQ2 How does the power consumption of stereo vision systems (RealSense D435, ZED, eCapture G53) compare under similar flight-like conditions?
• RQ3 What is the energy efficiency of event-based cameras (DAVIS346 DVS) relative to conventional stereo cameras in dynamic and static scenes?
• RQ4 How does scene activity (e.g., moving vs. static objects) affect the power draw of event cameras during simulated flapping-wing motion?
• RQ5 Can event cameras significantly reduce onboard perception energy use, thereby extending flight endurance in autonomous flapping-wing robots?

Key findings

• The launching and ascending phases together account for 52.9% of total energy consumption, while cruising flight consumes 47.1%.
• The DAVIS346 event camera consumes less than 200 mW, significantly lower than the 300–700 mW range observed in conventional stereo cameras.
• Event rate in the DAVIS346 DVS directly correlates with power consumption, with higher motion in scenes increasing energy draw, while static scenes trigger minimal events and negligible power use.
• The eCapture G53 stereo camera showed the lowest power draw among conventional stereo systems (300–400 mW), but still exceeded the event camera’s efficiency.
• The RealSense D435 and ZED cameras consumed 500–700 mW, making them less suitable for long-duration missions due to high energy demands.
• The dependency of event camera power on scene activity enables adaptive energy use—only relevant visual changes trigger consumption, making them ideal for long gliding phases.