Robotics
Work description
Long-range underwater navigation based on seabed morphology and infrastructure geometry in unstructured environments The doctoral research will focus on the development of advanced methodologies for long-range and long-endurance autonomous underwater navigation, exploiting seabed morphology, infrastructure geometry, and other natural or artificial elements of the environment as references for localisation. The research will aim to reduce reliance on external positioning infrastructure and limit the accumulation of error inherent in inertial navigation over extended missions, particularly in GPS-denied scenarios and environments with limited availability of artificial references. The objective is to develop solutions capable of maintaining high levels of navigation accuracy and consistency over long trajectories by exploiting environmental information acquired throughout the mission. The work will consider the integration of navigation sensors with acoustic and, where appropriate, optical sensors, enabling persistent environmental features to be identified and reused, such as seabed morphological variations, offshore structures, cables, pipelines, foundations, harbour infrastructure, and other relevant geometries. These references will be exploited to provide periodic corrections to the position estimate and reduce the growth of uncertainty over time. A central component of the research will be the development of methods capable of operating over missions covering large spatial extents, where navigation accuracy depends on the ability to recognise previously observed references, associate observations with existing maps or maps constructed during the mission, and consistently manage accumulated uncertainty. The objectives of the work include: Developing methods for representing the underwater environment that are suitable for long-range navigation, using bathymetric maps, three-dimensional representations, point clouds, and geometric or semantic descriptors. Investigating robust methodologies for the detection, extraction, and association of environmental features, capable of operating under varying viewpoints, resolutions, and acquisition conditions. Developing terrain- and infrastructure-based localisation algorithms that relate observations acquired by the vehicle to maps available beforehand or constructed during the mission itself. Studying strategies for periodic correction of navigation error through the recognition of previously observed areas, matching of persistent features, and identification of suitable environmental references. Developing sensor-fusion and state-estimation methodologies, with particular emphasis on the propagation and management of uncertainty during long-duration missions. Investigating criteria for selecting the most informative environmental references, prioritising observations that maximise the reduction of navigation uncertainty over long trajectories. Exploring, where relevant, information obtained from other robotic systems or mobile infrastructures as an additional source of navigation correction or validation, without assuming permanently multi-robotic operation. Evaluating the performance of the developed methodologies in long-range missions, considering accumulated positioning accuracy, robustness, navigation-solution availability, computational requirements, and real-time operation. Implementing and experimentally validating the developed solutions through a progressive evaluation process involving simulation, trials in controlled environments, and field campaigns using underwater robotic platforms. The work is expected to contribute to the development of underwater systems capable of undertaking increasingly long and autonomous missions by using the environment itself as a source of information for controlling navigation error and reducing reliance on external positioning infrastructure.
Academic Qualifications
Hold a Master’s degree in Electrical and Computer Engineering,?or any other field?that the jury considers aligned with?the call.
Minimum profile required
Hold a Master’s degree in Electrical and Computer Engineering,?or any other field?that the jury considers aligned with?the call, and have a scientific and professional CV that demonstrates a suitable profile for the activities to be carried out.Not have previously been awarded a doctoral grant or a doctoral grant in companies directly funded by FCT, regardless of its duration.Reside permanently and habitually in Portugal at the start date of the period of the work plan to be carried out abroad. This requirement applies to both Portuguese and foreign citizens.Not hold a doctoral degree.
Preference factors
The candidate's experience will be especially valued in one of the following research lines: experience in underwater robotics.
Application Period
Since 01 Oct 2026 to 15 Oct 2026
Centre
Robotics and Autonomous Systems
Scientific Advisor
Eduardo Silva