Autonomous spacecraft landing through human pre-attentive vision.

Area of Science:

• Robotics and Autonomous Systems
• Computational Neuroscience
• Space Exploration Engineering

Background:

• Autonomous spacecraft landing requires advanced hazard avoidance and landmark detection.
• Current methods often rely on pre-existing maps and scientific criteria.
• Future space missions demand greater spacecraft autonomy for unpredictable environments.

Purpose of the Study:

• To develop a bio-inspired computational model for autonomous spacecraft descent.
• To leverage human pre-attentive vision principles for spacecraft guidance.
• To achieve safe and fuel-efficient landings on extraterrestrial bodies.

Main Methods:

• Utilized a computational model of human pre-attentive vision.
• Integrated spacecraft dynamics and optimality principles for minimum fuel consumption.
• Employed salient target detection for trajectory retargeting under safety constraints.

Main Results:

• The developed algorithm's retargeting decisions were indistinguishable from human subjects.
• The system matched human performance in terms of areas, occurrence, and timing of retargeting.
• Demonstrated successful guidance of spacecraft descent using bio-inspired vision.

Conclusions:

• Bio-inspired approaches can effectively guide autonomous spacecraft landings.
• Human pre-attentive vision provides a robust framework for decision-making in unknown environments.
• The developed algorithm offers a promising solution for enhancing spacecraft autonomy in space missions.