The aim of my Ph.D. research was to build a robotic artifact mimicking biological functions identified in foraging insects like desert ants. These functions comprise the acquisition and processing of visual information, and relative navigation, well-known as the insects' path integrator. All together, these functions allow a six-legged walking robot, called AntBot (see the video below), to autonomously navigate in an unknown outdoor environment, just like desert ants Cataglyphis do. Autonomous navigation without GPS and localization were at the core of the of my research. As a Bioroboticist, my starting point was to figure out that this technological challenge is already solved by desert ants which are able to keep tracking their position with respect to the nest entrance while foraging for food over several hundred meters, and then return to the nest following a straight-forward path. This outstanding behavior has been reproduced on-board the AntBot robot, thus confirming hypotheses formulated years ago by entomologists, while providing us with minimalist, yet robust, outdoor-autonomous navigation system that could serve as a backup system in case of failure with conventional systems.

Robotics implementation required transposition of these biological functions into mathematical equations and electronic realizations. In that respect, a novel celestial compass was developed to mimic the dorsal rim area (DRA) of insects' compound eyes. Photoreceptors in the DRA are known to be sensitive to the polarization state of the skylight in the ultraviolet (UV) range, and are used to estimate the insect's orientation. AntBot's celestial compass is composed of two UV-sensitive photodiodes topped with rotating linear polarizers. A novel algorithm was proposed to estimate the robot's heading with a final accuracy of 0.4°, with strong resilience towards the environmental conditions (UV index, cloud cover, canopies). Furthermore, AntBot was endowed with a mathematical model for the desert ants' path integrator. This model is highly computationally efficient and is therefore biologically plausible. Outdoor experiments revealed AntBot's ability to reach centimeter accuracy. These outstanding results are meant to drive entomologists' attention on the high efficiency of a navigation strategy solely based on path integration.

The general purpose of Biorobotics is twofold: (i) to assess the operational aspect of the knowledge extracted from the observation of the living world by means of robotic implementation and testing in the same environmental conditions as its living model, and (ii) emerge new solution techniques for seemingly complex technological challenges. This approach allows to investigate on biological models beyond simulations, by exposing robots to natural hazard (light changes, temperature, cloud cover). At the crossroads between biology and robotics, the AntBot robot is the very first open-source hexapod platform equipped with minimalist visual sensors mimicking the compound eye of insects, and dedicated to testing biological models in real environments.

VivaTechnology - Paris May 16-18, 2019

AntBot has been exhibited at the VivaTechnology exhibition in Paris from May 16 to 18, 2019, at EDF's invitation (French electric utility company). Demos were performed during these three days on the Robot Park, including interviews with the French Minister of the Armed Forces Florence Parly, French Minister of the Economy and Finance Bruno Le Maire, and the French Secretary of State for the Digital Economy Cédric O.

Publications (selected)

[11] Insect-inspired omnidirectional vision for autonomous localization on-board a hexapod robot.
Dupeyroux, J., Lapalus, S., Brodoline, I., Viollet, S., & Serres, J. R. (2020, September).
In 28th Mediterranean Conference on Control and Automation MED 2020 (p. 893-898). IEEE.
[10] Bio-inspired celestial compass yields new opportunities for urban localization.
Dupeyroux, J., Viollet, S., & Serres, J. R. (2020, September).
In 28th Mediterranean Conference on Control and Automation MED 2020 (p. 881-886). IEEE.
[9] Dispositif de détection du cap d'un véhicule par détection de photons polarisés linéairement.
Dupeyroux, J., Monnoyer, J., Serres, J., & Viollet, S. (2020).
Patent no. FR3086088.
[8] AntBot: A six-legged walking robot able to home like desert ants in outdoor environments.
Dupeyroux, J., Serres, J. R., & Viollet, S. (2019).
Science Robotics, 4(27), eaau0307.
[7] An ant-inspired celestial compass applied to autonomous outdoor robot navigation.
Dupeyroux, J., Viollet, S., & Serres, J. R. (2019).
Robotics and Autonomous Systems, 117, 40-56.
[6] Polarized skylight-based heading measurements: a bio-inspired approach.
Dupeyroux, J., Viollet, S., & Serres, J. R. (2019).
Journal of the Royal Society Interface, 16(150), 20180878.
[5] A hexapod walking robot mimicking navigation strategies of desert ants Cataglyphis.
Dupeyroux, J., Serres, J., & Viollet, S. (2018, July).
In Conference on Biomimetic and Biohybrid Systems (pp. 145-156). Springer, Cham.
[4] M²APix: a bio-inspired auto-adaptive visual sensor for robust ground height estimation.
Dupeyroux, J., Boutin, V., Serres, J. R., Perrinet, L. U., & Viollet, S. (2018, May).
In 2018 IEEE International Symposium on Circuits and Systems (ISCAS) (pp. 1-4). IEEE.
[3] A novel insect-inspired optical compass sensor for a hexapod walking robot.
Dupeyroux, J., Diperi, J., Boyron, M., Viollet, S., & Serres, J. (2017, September).
In 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (pp. 3439-3445). IEEE.
[2] A bio-inspired celestial compass applied to an ant-inspired robot for autonomous navigation.
Dupeyroux, J., Diperi, J., Boyron, M., Viollet, S., & Serres, J. (2017, September).
In 2017 IEEE European Conference on Mobile Robots (ECMR) (pp. 1-6). IEEE.
[1] Hexabot: a small 3D-printed six-legged walking robot designed for desert ant-like navigation tasks.
Dupeyroux, J., Passault, G., Ruffier, F., Viollet, S., & Serres, J. (2017, July).
In 2017 20th IFAC World Congress (pp. 1-4).