Portrait of Robert Fonod

Robert Fonod, PhD

Engineer & researcher working at the intersection of guidance, navigation & control, computer vision, and machine learning.

GNC and CV/ML Engineer, R&D at SenseFly, now EagleNXT · Lausanne, Switzerland

Email LinkedIn Google Scholar GitHub Hugging Face

About

I enjoy solving hard technical problems with practical impact, especially where data, algorithms, and dynamic systems meet. My work spans academia and industry: from spacecraft autonomy, air-defense guidance, and fault-tolerant control, through drone-based urban traffic monitoring, to real-time embedded AI on aerial platforms.

I hold a PhD in automatic control, signal & image processing from the University of Bordeaux and a Master of Computer Science in data science from the University of Illinois Urbana-Champaign. I like bridging model-based domain expertise with data-driven methods, and seeing the result run on real hardware.

45+ peer-reviewed publications 800+ citations · h-index 14 30+ talks & posters worldwide 300+ manuscripts reviewed

Experience

Research

Computer Vision & Machine Learning

Multiple object detection and tracking in drone imagery, georeferenced trajectory extraction, embedded AI on aerial edge hardware, and deep learning for traffic state estimation, forecasting, and vehicle re-identification.

Representative paper Code
All publications in this theme →

Guidance, Navigation & Control

Cooperative guidance and estimation, ballistic 3D guidance design, nonlinear state Kalman and particle filtering in non-Gaussian (Cauchy) environments, and vision-based navigation and pose estimation for autonomous aerospace vehicles.

Representative paper Representative paper
All publications in this theme →

Fault Detection, Isolation & Recovery

Model-based robust fault detection and isolation, fault-tolerant control and control re-allocation for safety-critical systems, applied to spacecraft rendezvous in the Mars Sample Return mission context.

Representative paper
All publications in this theme →

Research conducted within projects funded by ESA, the U.S. Air Force, NSF, SNSF, Innosuisse, the ETH Board, and the U.S.–Israel Binational Science Foundation, in collaboration with industry partners including Thales Alenia Space and Airbus Defence & Space.

Full list on the publications page or on Google Scholar.

Software & Data

geo-traxPython · PyTorch · pip install geo-trax

End-to-end framework for extracting stabilized, georeferenced vehicle trajectories and kinematics from drone video; the pipeline behind the Songdo datasets.

GitHub PyPI Paper arXiv Models Live demo Video Zenodo Project site

stabiloPython · pip install stabilo

Video and object-trajectory stabilization library built on feature matching and robust homography estimation with a custom reference-frame strategy; core component of geo-trax.

GitHub PyPI Paper arXiv Video Zenodo

stabilo-optimizePython

Companion tool for stabilo: benchmarks and tunes stabilization accuracy against runtime performance via hyperparameter optimization.

GitHub Paper arXiv Zenodo

hbb2obbPython · PyTorch · pip install hbb2obb

SAM-assisted conversion of horizontal (axis-aligned) bounding-box annotations into oriented bounding boxes, with ensemble methods and evaluation tools.

GitHub PyPI Zenodo

deepsleep2Python · PyTorch

Compact U-Net-style CNN (~740K parameters) for millisecond-level sleep arousal segmentation from 13-channel polysomnographic recordings.

GitHub Paper Zenodo

songdo-trafficDataset · CC BY 4.0

About 700,000 georeferenced vehicle trajectories from 20 intersections in Songdo, South Korea, captured by a swarm of 10 drones at ~30 Hz: positions, speeds, accelerations, dimensions, and lane assignments.

Hugging Face Zenodo Paper arXiv Code

songdo-visionDataset · CC BY 4.0

5,419 annotated 4K drone frames with ~272,000 vehicle instances (car, bus, truck, motorcycle) in COCO, YOLO, and Pascal VOC formats; companion detection benchmark to songdo-traffic.

Hugging Face Zenodo Paper arXiv Trained model

More projects on GitHub and Hugging Face →

Awards & Recognition

Fellowships: Technion Postdoctoral Fellowship (2015–2017) · ESA Networking Partnering Initiative Fellowship (2011–2014) · Collegium Talentum Fellowship, Hungarian Academy of Sciences (2010–2013) · OeAD Fellowship, Austrian Agency for International Cooperation in Education and Research (2010)

Service