Exit Menu

Flaw detection and analysis in manufacturing environments using multi-sensory data fusion, machine learning and numerical computation

Applications are invited for ONE full-time PhD scholarship funded by Brunel University London. The studentship is for a period of 3 years, starting April 1st 2021 or at a mutually agreed time after that. The successful applicant will receive an annual stipend of £17,285 plus payment of their full-time home tuition fees. 

In manufacturing settings, structural flaws or malfunctions and operational inaccuracies are likely to happen, hindering the manufacturing process and impairing the quality of the production line and the structural integrity of the end product. Flaw detection and analysis becomes of paramount importance to help identify flaws at an early stage and empower the automatic decision-making required to rectify structural/operational errors and ensure a zero-breakdown optimal quality production.

This project will examine the acquisition of sensory data using a range of sensors such as RGG, iTOF, mmwave and SAR/Lidar used in a manufacturing setting. The project will then develop innovative methods for the fusion of these sensory modalities into one image that enables the 3D reconstruction/scanning of the manufactured products. Then, the project will look into image analytics and develop an algorithm for the low-level analysis of the generated 3D image with view of generating specific visual cues which represent a set of features.

This doctoral research will then design, build and develop a deep learning algorithm combined with numerical simulation algorithms to characterise flaws, e.g. cracks, in the manufactured products which will allow a trained neural network (supported by numerical computational models) to analyse and classify the type of structural flaws detected and support the decision-making required to eliminate/rectify it. An inverse numerical problem will be formulated using Artificial Neural Networks (ANN) for characterisation of the shape of the induced cracks in materials.

Using this data, finite element models of the cracks in the product will be developed to predict the path of the propagation of the cracks through the material during its service loading. This will allow computation of the induced damage in the product and probability of its failure under various loading scenarios.


  • Students must be eligible for home tuition fees
  • Experience in programming & simulation.
  • Demonstrable record of successful research project experience.
  • Basic knowledge of, and project experience with, Sensors and sensory data analysis, deep learning, Numerical Computing.

How to apply

Please email your application comprising all of the documents listed below by Noon on 12 February, 2021 to cedps-pgr-office@brunel.ac.uk and abdul.sadka@brunel.ac.uk. 

  • Your up-to-date CV;
  • A one A4 page personal statement setting out why you are a suitable candidate (i.e. your skills and experience);
  • Your degree transcripts and/or certificates;
  • Evidence of English language skills to IELTS 6.5 (or equivalent), if needed;
  • Name and contact details for two academic referees.