Embedded Files
I am passionate about deep-technology and how it can help society moves towards sustainability. I have dedicated the first decade of my career to oceans and I am presenting my contributions on this website.
I have worked for the first 5 years in offshore wind on aerodynamic design, loads and control as well as numerical tools development and validation. I have then worked for 5 years in subsea robotics on swarm control, deep learning and free-space optical communication.
I have a PhD in Engineering from UC Berkeley and two Masters (MSc), one from UC Berkeley and one from Ecole Polytechnique, France.
Hybrid Unmanned Underwater Vehicles
Hybrid Unmanned Underwater Vehicles
Commercial underwater robotic vehicles can be categorized into Remotely Operated Vehicles connected to a support vessel with a tether or Autonomous Underwater Vehicles pre-programmed to fulfill a mission.
These two extreme levels of autonomy limit the spectrum of missions that Unmanned Underwater Vehicles can fulfill. Hybrid vehicles with wireless optical communication can fill this gap and open up many new applications.
Swarm Control
Swarm Control
Do you know that more than 80% of the oceans are unexplored? We actually know more about the Moon and even about Mars than we know about our oceans!
This results from the heavy absorption of electromagnetic waves which cannot reach the ocean depths. Cellphones, Wi-Fi or GPS don't work underwater. Only optical signals can travel a meaningful distance.
To reach the deep sea, a guidance system combining Model Predictive Control and graph theory can steer a swarm of Unmanned Underwater Vehicles to amplify and relay the signal.
Transformer
Transformer
Transformers have drawn a lot of interests in the deep learning community for their performance in language modeling (ChatGPT...) but they can also be used for any type of data sequence!
Path planning and control of Unmanned Underwater Vehicles require prediction of ocean currents but today's forecast models are too computationally expensive to run on embedded systems.
Neural networks, such as the Transformer, enable to perform real-time in-situ prediction of ocean currents at any location in the world that can then be fed to the vehicle's control systems.
Free-Space Optical Communication
Free-Space Optical Communication
Free-space optical (FSO) communication is a promising technology to improve underwater wireless data communication which suffers from the low bandwidth and high latency of acoustic modems. LED modems are commercially available and offer high-bandwidth and low-latency over a moderate range.
The coherent light of Lasers can also be used to transmit data with several key advantages: very-high bandwidth, higher range, less impact of ambiant light and accurate positioning. A Pointing, Acquisition and Tracking system is necessary to reach the pointing precision required to sustain the communication link.
Twinfloat
Twinfloat
Offshore wind is a booming market with an expected installed capacity of 75 GW worldwide by 2020. Fixed foundations offshore wind turbines are installed in shallow waters up to 50-60m whereas the mean depth of the Atlantic ocean is 4km!
For higher depths, floating foundations are required to produce competitive electric power. The Twinfloat is a 5MW Vertical-Axis Wind Turbine (VAWT) leveraging multiple advantages of counter-rotating rotors to decrease the cost of electricity.
Pharwen
Pharwen
To develop new wind turbines, realistic simulations of offshore wind turbines operations are required to compute the design loads and power. However, on floating foundations, wind turbines experience unsteady aerodynamics due to the foundation's motions making State-of-the-Art models obsolete.
Pharwen is an aero-servo-elastic model developed to take into account the impact of these motions as well as the complex aerodynamics and aero-elasticity of Vertical-Axis Wind Turbines (VAWTs).
Individual Pitch Control
Individual Pitch Control
Vertical-Axis Wind Turbines (VAWTs) are seen as a promising technology for offshore wind deployment in deep sea markets. They have a lower center of gravity compared to Horizontal-Axis Wind Turbines, allowing to reduce the dimensions and thus the cost of the substructure. Their electrical components are also easier to reach simplifying the Operations and Maintenance activities.
They however suffer from lower performance compared to traditional Horizontal-Axis Wind Turbines (HAWTs) reducing their economic competitiveness. Blades Individual Pitch Control increases performance of VAWTs enabling them to close the gap with HAWTs and therefore making them a credible alternative for industrial deployments.
Page updated
Google Sites
Report abuse