PhD Candidate in multi-sensor fusion and representation learning for underwater robotics

Role Overview
This PhD project focuses on developing novel methods for multi-sensor fusion and representation learning that exploit remote sensing priors—satellite optical and multispectral imagery, synthetic aperture radar (SAR), aerial photogrammetry, and prior bathymetry—to augment what an underwater robot can perceive, localize within, and reason about. The aim is to advance principled cross-modal and cross-scale representation learning for the underwater domain, and to demonstrate that richer, uncertainty-aware environmental models lead to more capable and efficient autonomous underwater surveys.

The project will develop principled methods for fusing heterogeneous observation streams into coherent, uncertainty-aware environmental representations that an underwater robot can exploit for perception, habitat mapping, and adaptive survey planning. The central scientific challenge is cross-modal and cross-scale representation learning: how to encode geometric, acoustic, optical and environmental observations made at different scales, resolutions, modalities and times into a shared representation that is actionable for an onboard robot. Applications include benthic habitat mapping (cold-water coral reefs, kelp forests, seafloor geology), infrastructure inspection and persistent environmental monitoring. The project will investigate whether and how recent foundation models for earth observation and underwater perception can be adapted and extended to support this multi-modal, cross-scale fusion challenge.

The PhD candidate will be supervised by Professor Oscar Pizarro at the Department of Marine Technology (IMT), and the position is part of the recently funded Norwegian Centre for Embodied AI (NCEI). NCEI is one of Norway's six national AI centers, recruiting outstanding researchers to advance a universal science of embodied intelligence.

Responsibilities
- Complete your doctoral education leading to the PhD degree.
- Conduct and publish research of high quality within the framework described above.
- Develop fundamental contributions in multi-sensor fusion and representation learning for underwater robotics, with an emphasis on exploiting remote sensing priors.
- Develop implementable methods for onboard robot perception, localization, and adaptive survey planning.
- Conduct experimental deployments and field evaluation using NTNU's Fjordlab AUVs and other infrastructure with acoustic (MBES, SAS), visual, and hyperspectral sensing payloads.
- Participate in international activities such as conferences and/or research stays abroad.
- Collaborate with other researchers within the department, and across departments at NTNU.
- Supervise master's thesis students related to the project.
- Be prepared for changes to your work duties after employment.

Requirements and Qualifications
Required selection criteria:
- You must have a relevant master's degree in marine technology, cybernetics, mechatronics, control systems, computer science or equivalent, with strong training in robotics, computer vision, state estimation, machine learning or statistical signal processing.
- Your course of study must correspond to a five-year Norwegian course, where 120 credits have been obtained at master's level. Master students graduating summer 2026 are eligible to apply.
- You must have a strong academic background from your previous studies and have an average grade from your Master's degree study, or equivalent education, which is equal to B or better compared to NTNU's grading scale. If you do not have letter grades from previous studies, you must have an equally good academic foundation. If you have a weaker grade background, you may be considered if you can document that you are particularly suitable for a PhD education.
- You must meet the requirements for admission to the faculty’s Doctoral Programme in Engineering.
- Good oral and written presentation skills in English.

Preferred selection criteria:
- Solid theoretical background in robot perception, navigation and mapping.
- Background in one or more of: sensor fusion, probabilistic state estimation, underwater perception, remote sensing image analysis, or 3D reconstruction.
- Deep understanding of modern machine learning including foundation models and/or self-supervised representation learning.
- Solid programming skills in Python and C++.
- Experience with ROS2 is a plus.
- Experience with simulators for robotics, geospatial data tools and remote sensing processing.
- Strong skills in mathematics, excellent capacity for mathematical formalism, and ability to grasp new concepts quickly.
- Real-world deployment of robotic systems and fieldwork experience, preferably at sea.

Personal characteristics:
- Ability to work independently in a structured way, set goals and make plans to achieve them.
- Scientifically curious, critical, creative and open to new research challenges.
- Ability to incorporate feedback, resourcefulness and persistence in solving technical problems.
- Reliable, with the ability to work effectively and proactively as part of a multidisciplinary team.
- Ability to communicate clearly in person and in writing.
- Ability to present and discuss your research with other professionals.
- Ability to work constructively under pressure or in the face of adversity.
- Flexible and open to adjusting the plan for the project as needed.

What They Offer
- An exciting job with an important mission in society.
- Developing tasks in a strong and international professional environment.
- Career guidance and follow-up during the PhD period.
- Open and inclusive working environment with committed colleagues.
- Favorable terms as a member of the Norwegian Public Service Pension Fund (SPK).
- Access to employee benefits as a PhD Candidate at NTNU.
- Gross salary of normally NOK 550,800 per annum depending on qualifications and seniority.
- Employment period of 3 years.