Egomotion estimation with optic flow and air velocity sensors

Area of Science:

• Robotics and Autonomous Systems
• Bio-inspired Navigation
• Sensory Processing in Flying Organisms

Background:

• Flying insects and aerial vehicles require accurate self-motion and environmental perception for navigation.
• Existing methods often rely on external references like GPS, which are not always available.
• Insects utilize optic flow and wind-sensitive organs for flight control and orientation.

Purpose of the Study:

• To develop a novel algorithm for estimating egomotion, wind velocity, ground slope, and flight height using onboard sensors.
• To demonstrate the feasibility of insect-like absolute egomotion estimation for autonomous navigation.
• To provide a robust navigation solution for aerial vehicles in GPS-denied environments.

Main Methods:

• Development of the 'opto-aeronautic algorithm' integrating optic flow and air velocity sensor data.
• Testing the algorithm using recorded moth flight data during pheromone plume tracking.
• Validation across diverse wind and ground slope conditions.

Main Results:

• The algorithm accurately estimates egomotion, wind, ground slope, and flight height.
• Achieved a mean absolute error of 7% or less for height estimation across all tested conditions.
• Demonstrated the capability for insects to determine absolute egomotion (groundspeed) and wind direction.

Conclusions:

• The opto-aeronautic algorithm offers a viable method for egomotion and environmental estimation using limited onboard sensors.
• This bio-inspired approach has significant implications for autonomous aerial vehicle navigation, especially in GPS-unavailable settings.
• The findings support the hypothesis that insects can determine their absolute egomotion using available sensory inputs.