Open Positions

Open positions, when available, will be announced below. However, if you are a highly motivated and capable student interested in gaining research experience and co-authoring high-quality publications, you are welcome to get in touch. We occasionally offer voluntary research internships, which are not salaried but provide valuable hands-on experience. To express your interest, please email your CV to oktay.cetinkaya@newcastle.ac.uk. 

Closed Positions

A funded internship position is available (NCL STUDENTS ONLY)! 

I am offering an exciting internship opportunity titled "[G8WAI] AI-driven Desktop Surveys for Low-level LoRaWAN Network Design" for one student at Stage 2 or above (including PGRs), studying Electrical and Electronics Engineering, Computer Science, or a related discipline at the School of Engineering, Newcastle University.

This internship is expected to be for a total of 60 hours. You will work up to 5 hours per week between March and May 2026. Working days and times are flexible, and the work will take place on campus, likely on Merz Court (ISC Research Group PhD offices or the Sensors Laboratory).

To apply, please visit https://mycareer.ncl.ac.uk/leap/jobs.html?id=100554&service=Careers%20Service 

About the role

LoRaWAN has become a key infrastructure layer for smart-city applications, such as air-quality monitoring, waste management, and smart metering, and even medium-sized urban areas often require dozens of gateways to achieve reliable coverage. As these deployments scale, careful gateway placement becomes increasingly important, which begins with a low-level network design once high-level simulations identify a set of candidate streetlights that can host gateways. A substantial amount of time is then spent on desktop surveys, where network planners manually inspect each candidate streetlight using Google Street View to determine whether it is suitable for deployment. This step is essential because councils impose strict rules on where equipment can be mounted, and network performance can be heavily affected by the immediate surroundings. To be specific, streetlights carrying large traffic or information signs, festive decorations, or additional mounted infrastructure are generally not permitted. From a network perspective, streetlights surrounded by dense foliage, overhanging branches, or obstructing buildings are avoided because they limit propagation paths and may lead to localised coverage shadows. Tofa Ltd have already developed a Python tool that automatically retrieves Street View images for selected streetlights using their coordinates. The script computes the correct viewing angle by selecting a nearby point on the road and calculating the bearing towards the streetlight, ensuring that each screenshot shows the intended structure rather than the generic view usually presented when searching coordinates directly on Google Maps. However, while this significantly reduces the time spent navigating Street View, it only generates images and does not support candidate confirmation or decision-making. Using this script, the previous low-level design projects have produced labelled datasets of “approved” and “rejected” streetlights, which offer a strong foundation for developing a supervised machine learning model that can support and accelerate desktop surveys. 

The main aim of this internship is to develop, calibrate, and evaluate a machine learning model that predicts whether a candidate streetlight is suitable for gateway installation, and to demonstrate its use within a simple web-based tool/dashboard (if time allows). You will be closely supervised by Dr Oktay Cetinkaya (School of Engineering) in support of Mr Omer Ari (Director of Network Planning & Optimisation, Tofa Ltd), and during the project, you will have direct exposure to a real deployment workflow and the chance to contribute to an emerging AI-assisted planning tool for low-level LoRaWAN network design.

Key tasks

• Examine and prepare the labelled dataset of approved and rejected streetlights (using the previous work)

• Implement initial baseline approaches using classical computer vision features and simple classifiers to establish early benchmarks (identify common rejection patterns, e.g., signage coverage, festive decorations, camera or sensor units, vegetation density, surrounding trees, nearby buildings, and private roads)

• Develop a transfer-learning model based on a pre-trained convolutional neural network

• Fine-tune the model using the curated dataset and evaluate performance using accuracy, precision, recall, and confusion-matrix analysis

• Calibrate predicted probabilities and define a practical decision-making policy (e.g., auto-accept/reject above 0.9 confidence; human review for 0.6-0.9; manual inspection below 0.6). 

• (If time allows) Integrate the image-capture script and classifier results into a simple web-based tool

• (If time allows) Build a workflow allowing users to upload an Excel file of streetlight coordinates, generate images, display model decisions with confidence scores, apply manual corrections, and export a final validated list

Essential skills

• Good experience with Python 

• An interest in machine learning/AI

Desirable skills

• Critical thinking and problem-solving

• Creativity and innovation in approaching challenges *Strong communication and presentation skills 

• Time management and organisational ability 

• Willingness to learn new tools and methods

Degree discipline

Students studying Electrical and Electronics Engineering, Computer Science, or a related discipline are welcome to apply.

To be eligible to apply, you must be a Newcastle University undergraduate/postgraduate student registered on a programme of study throughout the 2025-26 academic year. For full details of eligibility, see the Working on Campus website.

Work done as part of this role can count towards an NCL+ Award, which can help you reflect on the skills you have gained and how to use them when applying for jobs in the future; to find out more, visit our webpages. 

A funded internship position is available (NCL STUDENTS ONLY)! 

I am offering an exciting internship opportunity titled "Mind the Gap!: Reconstructing Missing Water Meter Data with Machine Learning" for one student at Stage 2 or above (including PGRs), studying Electrical and Electronics Engineering, Computer Science, or a related discipline at the School of Engineering, Newcastle University.

This internship is expected to be for a total of 60 hours. You will work up to 5 hours per week between March and May 2026. Working days and times are flexible, and the work will take place on campus, likely on Merz Court (ISC Research Group PhD offices or the Sensors Laboratory).

To apply, please visit https://mycareer.ncl.ac.uk/user/app/jobs/view/ZXO47 

About the role

In the UK, ageing water pipelines lose over three billion litres of treated water every day due to leaks that are hard to detect. Replacing entire networks is neither practical nor affordable, so water companies are turning to digital monitoring. Smart meters, combined with flow and pressure sensors, are now being deployed to track hourly consumption data, which can be interrogated to spot anomalies that may pinpoint leaks or bursts. However, these meters face a major challenge: data gaps. The UK Water Services Regulation Authority (Ofwat) requires an 85% hourly data success rate (DSR), yet current wireless technologies such as LoRaWAN and NB-IoT often deliver only 60-70%, even in good signal conditions. Closing this gap by adding more gateways, i.e., densifying the network to minimise packet losses, would cost water companies millions, involving also lengthy planning permissions and public reluctance. This project aims to address this gap using data science and machine learning to reconstruct missing water meter data, improving network reliability and leak detection without costly infrastructure expansion.

Working with a large dataset provided by Connexin Ltd, as an intern, you will analyse millions of historical meter transmissions to identify patterns in household consumption and develop models to reconstruct missing readings. Methods such as recurrent neural networks (RNNs), long short-term memory (LSTM) networks, or other modern architectures (e.g. Transformer-based or GANs) may be explored, depending on progress. You will be closely supervised by Dr Oktay Cetinkaya (School of Engineering) and supported by Mr Jarod Gillespie (Head of Strategy and Delivery, Connexin), ensuring you receive ongoing guidance and feedback throughout the project. This role provides hands-on experience at the intersection of IoT, machine learning, and sustainability, developing practical skills in data analysis and model validation while contributing to one of the UK’s key infrastructure challenges.

Key tasks

• Preprocessing and cleaning raw smart-meter data to locate gaps (Week 1-2)

• Training predictive models that learn from complete sequences of data (Week 3-4)

• Testing model accuracy by deliberately removing known readings and checking how well the model predicts them, using measures like RMSE and MAE (Week 5)

• Benchmarking against simple alternatives, such as linear interpolation (Week 5)

• Compiling results, preparing the final report, and conducting the industrial supervisor feedback session (Week 6)

Essential skills

• Basic experience of Python or MATLAB

• An interest in data analysis, IoT, distributed sensing, and machine learning/AI

Desirable skills

• Critical thinking and problem-solving

• Creativity and innovation in approaching challenges

• Strong communication and presentation skills

• Time management and organisational ability

• Willingness to learn new tools and methods

Degree discipline

Students studying Electrical and Electronics Engineering, Computer Science, or a related discipline are welcome to apply.

To be eligible to apply, you must be a Newcastle University undergraduate/postgraduate student registered on a programme of study throughout the 2025-26 academic year. For full details of eligibility, see the Working on Campus website.

Work done as part of this role can count towards an NCL+ Award, which can help you reflect on the skills you have gained and how to use them when applying for jobs in the future; to find out more, visit our webpages. 

Research Assistant/Associate in Sensor Systems

I'M HIRING! Join me at Newcastle University to develop networked sensor systems, powered by ML algorithms, to monitor and assess food freshness, helping reduce household food waste.

Fixed-term: 9 months.

Deadline: 12th January 2025.

Details: https://www.linkedin.com/jobs/view/4104067358/