Polarized skylight | Stéphane Viollet

Navigating in Global Navigation Satellite System-denied or unmapped environments will, over the coming decade, become one of the 10 biggest challenges in robotics. Currently, autonomous robots rely on Global Navigation Satellite System (GNSS), Inertial Navigation Systems (INS) and ground-based antennas to triangulate or correct GNSS signals (5G networks or Real-Time Kinematic (RTK) networks), astronomical navigation, gyrocompass navigation, and vision-based or lidar-based SLAM. However, as summarized in (Serres et al., 2024), many animals are able to navigate or migrate over extremely long distances without using localization techniques developed by humans. For instance, desert ants use a powerful navigational tool termed optical path integration to locate their nest. When returning, desert ants follow the shortest possible route, a straight line, even in featureless or unfamiliar terrains. By integrating a directional compass and distance information from their vision, desert ants calculate a vector from their visual inputs, and this leads them home (see R. Wehner’s book ).

The aim of this project is precisely to study new methods based on the processing of the polarized skylight to determine in real time the heading to the true North but also the location on earth (latitude and longitude) by means of only polarimetric information. We developped a fully open access and open source simulator, named OpenSky (Moutenet et al., 2024) providing realistic images from a simulated polarimetric camera oriented toward the sky. We recently developed a method named SkyPole (Kronland-Martinet et al., 2023) able to locate the celestial pole (pole point) from the pattern of degree of polarization of the sky measured by a polarimetric camera. Tje SkyPole method was recently improved by including the knwoledge of the solar declination in the estimation (Kronland-Martinet et al., 2025). We also published preliminar results about the polarized sklylight analysis in the UV range (Moutenet et al., 2025) including the development of a new system (Ravolafidy et al., 2026) combining two polarized cameras (UV and visible ranges).

(Left) Global view Credit: Léo Poughon) and mechanical layout of the measurement device used to collect polarimetric images during two months (Poughon et al., 2023) and (Poughon et al., 2024). A fully anotated database composed of several thousand of images is available here including Python code to make it user friendly. (Middle) OpenSky simulate the measurement of sky polarization properties, i.e., the light intensity, the angle of polarization and the degree of linear polarization of light seen through a polarimetric camera. OpenSky is an open source simulator available here. (Right) The green circle is the radial invariance circle, the red circle is the plane invariance circle computed from analytical calculations (see CITE). The colored half sphere is the simulated absolute difference between two DoLP patterns associated with the Sun’s positions S1 and S2 at two different times. The red dot is the NCP, and the maximum value of the absolute difference of DoLP is in yellow. The NCP can be found at the intersection of the two invariances when the solar declination is null or by estimating the axis of rotation of the invariance circles resulting from the rotation of the sun.

References

2026

2025

2024

2023