Nathan Cusson-Nadeau

I am an engineer, roboticist, and Green-Tech enthusiast currently working on the research and development of aerial drones. I am primarily interested in the advancement of robotics technology from all facets, but have particular expertise in control theory, software development, and machine learning.

I graduated from UC San Diego for both my MS and BS in Mechanical Engineering and Environmental Engineering respectively and formerly served as a researcher in the Contextual Robotics Institute in the Safe Autonomous Systems Lab where I was advised by Professor Sylvia Herbert.

Across my career, I have worked on software development and controls for robotic systems, autopilot for Unmanned Aircraft Systems (UAS), building automation control systems (BACS), and environmental consulting.

Email  /  CV /  Github /  LinkedIn

Projects & Research

My professional work and research history spans the broad domains of advanced control theory, robotics software development, robot motion planning, state-estimation, green-technology, building automation control systems, and oceanography.

Demonstrated that refineCBF [arxiv] from Sander Tonkens and Sylvia Herbert can be applied on hardware-in-the-loop use case and in the presence of changing obstacles.

The algorithm provides a constructive method to take an initial candidate control barrier function (CBF) and refine it to be the least conservative and safe control barrier function upon convergence. CBF's can be used to enforce and encode safety constraints into a control system through the use of a safety filter. Thus, a robot with a goal-seeking control policy can maintain optimality while staying as safe as possible.

Implemented a novel variation of the A* algorithm from scratch, coined Delayed Weighted A*, which improves performance of algorithm in larger and more complicated configuration spaces.

Used the dynamic programming principle to construct the optimal control policy for an agent in a randomly generated door-key grid-world environment to reach a goal. Received highest grade in class for implementation and report.

Implemented particle filter SLAM from scratch to reconstruct a real autonomous car's environment using sensor fusion of LIDAR and IMU data.

Used logistic-regression to train a pixel-classifier to detect pixels of recycling bin blue. Then used this classifier to detect recycling bins in images using the Python OpenCV libary.

Used MATLAB to analyze oceanographic data from the Regional Ocean Modeling System (ROMS) to assess efficacy of model in predicting oceanic conditions such as temperature, and vorticity.

Work Experience

• Design, simulate, and verify precision actuation control and pointing algorithms for Small Unmanned Aerial Systems (SUAS) and other sensors as well as various actuators and mechanisms.
• Provide technical contributions to Guidance, Navigation, and Control (GNC) analysis tasks to include algorithm development for guidance and control law, navigation, and target state estimation.
• Create, edit, and debug software written in C/C++, Matlab, Python, and/or other languages. Develop, modify, maintain, and improve software simulations, tools, and scripts.

Developed regression tests in the Python and C programming languages for the NGPS system of the Gray Eagle Drone, which utilizes Kalman Filtering techniques to enable the autopilot system to land in GPS-denied scenarios. Required an intimate understanding of Kalman Filtering and the drone's autopilot software.

Designed and programmed building automation control systems for HVAC and lighting systems for clients in the greater Los Angeles area, such as NASA-JPl, UCLA, and NBC Universal.

Designed building automation control systems for HVAC systems for clients in the greater San Diego area, such as UCSD, Qualcomm, Apple, and One America Plaza. The most notable and rewarding projects I completed would be the Ultracold Strontium Lab's specialized HVAC control system, and the Nuevo West & East graduate housing campus at UCSD.

Performed air emission data analysis and modeling for environmental projects along with other miscellaneous enviromental matters. Fulfilled and processed various permit applications for air, water, and land use related projects.

Student researcher under the advisement of Esteemed Research Oceanographer Arthur J. Miller. Assessed the efficacy of the Regional Ocean Modeling System (ROMS) in predicting oceanic conditions such as temperature, and vorticity. Discovered the presence of the Butterfly Effect in the ROMS model.

Education

Certifications