-2014/2015: ISAE (Institut Supérieur de l’Aéronautique et de l’Espace), specializing in Embedded Systems, and Signal/Image Processing. Toulouse, France.

-2012/2013: École Normale Supérieure de Rennes (a highly ranked Research College in France) in the field of Mechatronics (dual Masters in Mechanics and Electronics). Rennes, France.

3rd year, specializing in Signal, Image, Communication and Multimedia. 2nd year in substitution at ENS Cachan. Noticeable course projects: programming on GPU (Nvidia cards), DSP and microcontroller; programming in C/C++ and VHDL. Grenoble, France.

Specific scientific undergraduate courses to prepare for the highly competitive admission exams to top French Engineering School. Centre International de Valbonne, France.

Designing the Inspector X4, a quadcopter for infrastructure inspection. It features a full frame sensor (Sony Alpha 7R) and weigh less than 2 Kg. The gimbal is designed to orient the camera upward without seeing the propellers, and being able to take photos from under the infrastructure. Designing the Power Riser, a tethered drone power system for sensitive site surveillance. It can supply drones with safe continuous energy, for up to 12 hours, at up to 150 meters. Main tasks : Mechanical and electrical design; realization of prototypes, 3D printing, electrical wiring, PCBs; assembly and testing; production launch. Software used : Solidworks (CAO), Eagle, Cambam, Cura, IDE Arduino, visual studio code.

I especially designed a quadcopter for aerial video. I used SolidWorks to create a simple and foldable design which can be easily machined. I also spent some time on minimizing vibrations and tuning the PID controller to get smooth videos. I got experience with the ArduPilot software to set up automatic flight modes (return to home in case of signal lost, follow a waypoint track..). I also worked on applications such as 3D reconstruction by photogrammetry using Agisoft PhotoScan. I was able to 3D reconstruct an house with an under 5 cm precision from GoPro shots.

Perception and estimation for dynamic interaction with a drone. The case study was “catching a ball thrown in the air with a drone” using only embedded sensors. I developed a stereovision camera and a software in order to estimate the 3D trajectory of the ball from the drone. Simulations performed using Matlab and software developed in C++ using OpenCV and ROS (Robot Operating System).

3D reconstruction of large scenes, using an RGB-D camera. I worked on a compact representation method of 3D point clouds, adapted to large scenes including free form objects. The method had to perform the reconstruction in real time, accurately with limited memory resources. Project developed in C++ using OpenCV and OpenGL libraries, but also g2o library (General Graph Optimization) and CGAL (Computational Geometry Algorithms Library).

Development of a C software on a microcontroller in order to perform real time acquisition on a quadrotor IMU (Inertial Measurement Unit). Comparison of two different IMUs, using ROS (Robot Operating System) for the data recording and Python for the data processing.

Participation in the CanSat contest supervised by the CNES (National Center for Space Studies). The objective was to create a satellite, in the volume of a can, which had to accomplish different missions such as atmospheric sounding during its parachute drop. My team was ranked 3rd and won the « Jury Favorites Prize ».

Ensuring receipts and shipments of engine parts in the military store (handling and logistics).