Dr A Munafo, Dr E Frajka-Williams, Prof Blair Thornton
No more applications being accepted
Funded PhD Project (European/UK Students Only)
About the Project
For long endurance/large area missions, novel navigation methods are needed to enhance the navigation abilities of Autonomous Underwater Vehicles (AUVs). Long-range underwater navigation is problematic due to the lack of external references (e.g. GPS) with typical solutions relying on dead-reckoning navigation or acoustic beacons deployed in fixed locations.
The limitations of these traditional methods include task interruption for surface navigation fixes, deployment of beacons at additional expense but with limited range, and additional navigational uncertainty in the case of vehicles which rely on bottom tracking for dead-reckoning. They also limit large-scale deployment in key areas such as under pack ice, ice-shelves or in Marginal Ice Zones.
As an alternative approach, this study will exploit the link between the navigation of the vehicles and the oceanographic data measured by the vehicle. In particular, using marine environmental information (e.g. temperature, salinity, or other parameters) with known maps of those properties, either from models or observations, will enable better localisation of the vehicle in the absence of other navigational aids. In addition, this link can be exploited for onboard updates of property maps, with the potential to enable better sampling of underwater oceanic temperature/salinity features.
As a first step, this study will provide a broad insight into the methodologies necessary to exploit ocean structures and ocean-based maps for robotic underwater localisation. The study will evaluate and quantify any increase in positional accuracy and mission lengths arising from different source data.
This PhD will combine robotics and ocean modelling and it is ideal for students with an interest in marine autonomous platforms. The student will use simulated conditions from several, high-resolution, realistic ocean models (1/12th degree global ocean model (including biogeochemistry) and a 1/60th degree Atlantic Margin Model which includes tides) and will develop 1) methodologies to assess and characterise the ability of state of the art ocean models to reproduce features that can be used for navigational purposes; 2) algorithms that are able to integrate environmental and/or biogeochemical ocean features into a formal localisation and navigational framework (e.g. Simultaneous localization and mapping, SLAM). Hypotheses will be tested against simulated and in-the-field data obtained through recent AUV campaigns.
The aim of the project will be to:
· Provide a detailed and accurate characterisation of the error that is required from the model to support a desired accuracy in the localisation of the robots.
Develop model-based navigation algorithms that fit within the more general concept of SLAM, where the vehicles construct or update a map of the environment while simultaneously keeping track of their own location within it.
The NEXUSS CDT provides state-of-the-art, highly experiential training in the application and development of cutting-edge Smart and Autonomous Observing Systems for the environmental sciences, alongside comprehensive personal and professional development. There will be extensive opportunities for students to expand their multi-disciplinary outlook through interactions with a wide network of academic, research and industrial / government / policy partners.
The student will be registered at University of Southampton, and hosted at National Oceanography Centre, Marine Autonomous and Robotic Systems.
The student will have access to resources at one of the top oceanographic research institutions in the world, including NOC’s large model datasets and to the Europe’s largest fleet of Marine Autonomous Systems, with opportunities to join trials and scientific surveys at sea using state-of-the-art robotic platforms.
Specific training will include:
· Use and deployment of autonomous systems
· AUV path planning and vehicle control
· Robot localization and navigation, including simultaneous localization and mapping (SLAM)
Environmental and biogeochemical ocean modeling.
Funding Notes
To be eligible for a full NEXUSS award (stipend and fees) a student must have:
No restrictions on how long they can stay in the UK
Been 'ordinarily resident' in the UK for 3 years prior to the start of the grant.
Not been residing in the UK wholly or mainly for the purpose of full-time education. (This does not apply to UK/EU nationals)
Potential PhD students are requested to apply using the University of Southampton postgraduate application form. For information on the application process and documents required please refer to the following webpage:
http://noc.ac.uk/education/gsnocs/how-apply
References
Lermusiaux, P F.j. "Adaptive Modeling, Adaptive Data Assimilation and Adaptive Sampling." Physica D: Nonlinear Phenomena 230.1-2 (2007): 172-96.
Lermusiaux, P F.j. et al. "Science of Autonomy: Time-Optimal Path Planning and Adaptive Sampling for Swarms of Ocean Vehicles." Springer Handbook of Ocean Engineering (2016): 481-98
Mahon, I., et al., "Efficient View-Based SLAM Using Visual Loop Closures." IEEE Transactions on Robotics 24.5 (2008): 1002-014