Understanding Underwater Coastal Soundscapes Using Machine Learning

Key information

Lead supervisor: Dr Tom Smith
Application deadline: ongoing
Project start date: 01 May 2026 or 01 October 2026
Project duration: 4 years (full-time)
Location: UCL Mechanical Engineering (Bloomsbury, London)
Studentship funding provided: Home tuition fees (currently £6,215/year) and maintenance stipend (currently £22,780/year) for 3.5 years

PhD project description

Background: 
Underwater acoustic recordings contain a wealth of data about marine traffic, levels of ecological activity, and even the weather. Being able to identify the presence of a vessel (either on the surface or below it) in a given area is important for defence and security applications and for protecting subsea infrastructure. Furthermore, understanding the soundscape of an area and how it changes over time is vital for conservation efforts and understanding how human activities and climate change are affecting the marine environment. 

Different marine species produce distinct calls (for example echo location clicks) which can be used to monitor population levels over time. Similarly, different marine vessels also have their own characteristic signatures depending on their size, propulsion configuration, and speed. In coastal environments, there is often a significant overlap between ecological and human activity, resulting in a complex soundscape that can be challenging to unravel. Machine learning techniques show a lot of promise for analysing these large complex acoustic datasets and identifying the makeup of a soundscape.

Aims: 
The aim of this project is to develop and use advanced data analysis and machine learning methods to determine the makeup of an underwater soundscape from hydrophone recordings. Building upon existing research into underwater soundscapes and machine learning methods for detecting different marine species, you will develop models that can analyse a soundscape over different time and frequency scales, allowing for different components to be identified and classified. A key focus will be on the identification of ship noise alongside other sources such as vocalisations from marine life. 

In addition to developing these models, you will have the opportunity to carry out field work to gather data for your research. The position also offers opportunities to support in teaching activities and work with other researchers with the Marine Research Group. As a PhD student at UCL, you will benefit from training in high-impact research. You will be encouraged to publish work in leading journals and present findings at national and international conferences.

Person specification

• Applicants are preferred to have a first-class undergraduate and/or master’s degree (or equivalent) in Mathematics, Physics, Engineering or Computer Science
• Applicants should have excellent analytical skills and a genuine interest in using them to understand complex problems
• Excellent organisational, interpersonal and communication skills, along with an interest in interdisciplinary research are essential
• Programming experience (e.g. Python, MATLAB) is highly desirable
• Previous experience in signal processing and/or machine learning would be advantageous
• Fluency and clarity in spoken English as well as good written English in accordance with UCL English requirements (TOEFL>92 or IELTS>6.5).

Eligibility

Please note that the available funding supports tuition fees at the Home/UK rate. Students who are eligible to pay fees at the UK rate are welcome to apply. Please refer to our website for further information about Home tuition fee eligibility.

International students who are eligible to pay tuition fees at the Overseas rate are also welcome to apply, however the tuition fees covered by the studentship will be limited to the Home/UK level. International students will be required to find additional funding for the remaining Overseas tuition fees.

Applicants whose first language is not English are required to meet UCL's English language entry requirements.

Please refer to this webpage for full eligibility criteria: Mechanical Engineering MPhil/PhD

How to apply

Eligible applicants should first contact Dr Tom Smith (tom.smith.17@ucl.ac.uk) to express their interest in the project. Please enclose the following documents:

1. A one-page statement outliningyour suitability for the project.
2. A two-page CV, include contact details for two referees.

After discussing the project with Dr Smith, eligible applicants should also submit a formal PhD application via the UCL website.

The supervisory team will arrange interviews for short-listed candidates.