New UROP opportunities will be listed here for one month, and thereafter will appear on the relevant faculty page (UROP website) until notified otherwise by the relevant member of academic staff.
Please note: The vast majority of advertised UROPs are advertised within academic departments for an internal audience (Imperial undergraduates) OR (and this is the route taken for most UROPs) are sourced by an individual student (whether an Imperial undergraduate OR an eligible undergraduate at another university) through direct contact with academic/research staff. Therefore this NEW OPPORTUNITIES page is normally used only by research groups who are looking to widen the audience for any UROP they are planning.
NEW (24 January 2024)
Exploring Potential Strategies for Robotic Structural Inspection and Monitoring
Corrosion is a prevalent degradation phenomenon that plagues many engineering structures, spanning vast surface areas and extended periods. The conventional approach involves either frequent monitoring at specific locations or infrequent scanning of larger areas, each with its own drawbacks. Recent technological advancements present an alternative approach: the deployment of resident robots allow sensors to manipulated robotically for 4D encoded measurements. This raises questions about the necessary number of robots and sensors for comprehensive area coverage and the potential extension/reduction of inspection intervals to gather insights into degradation rate variations. A recent study conducted by the Imperial NDE Group presents a potential solution to this question. The proposed UROP project will build on it and conduct a range of case studies that quantitatively describe the benefits of the proposed hybrid approach in various degradation scenarios.
The prospective UROP student will conduct case studies using a recently developed simulation code written in MATLAB. He/she will also have the opportunity to contribute to the code by introducing additional functionalities, improving computational efficiencies, and developing a graphical user interface (GUI). Findings from the simulation should be properly documented and presented regularly at group meetings. Positive results will be featured in journal publications that will be prepared in 2024. The candidate will be able to gain knowledge and experience in NDE, acoustics, statistics and programming.
Skills and experience required: Proficient in coding with MATLAB; Excellent reporting and presentation skills; Knowledge of statistics (Extreme value analysis, Monte Carlo simulation) is desirable.
Duration: Part-time during term time and/or full-time during the summer
Contacts: Dr Yifeng Zhang and Dr Frederic Cegla, Non-Destructive Evaluation (NDE) Group, Department of Mechanical Engineering, South Kensington campus. Email: firstname.lastname@example.org
NEW (December 2023)
Using Transformers (Masked Autoencoder) for Predicting high-dimensional Dynamical Systems with Irregular Time Steps
The accurate prediction of dynamical systems, especially those with irregular time steps and sparse observations, is a significant challenge in various fields like climate prediction, wildfire management, and fluid dynamics. Traditional models, such as CNNs or RNNs, often struggle with irregular temporal data and high-dimensional systems. This project proposes the development of a transformer-based model to address these challenges.
The primary objective of this project is to build and evaluate transformer models capable of predicting high-dimensional dynamical systems using sparse and irregular time observations. The project aims to leverage the unique strengths of transformers in handling irregular time steps and missing observations, which are limitations in traditional RNN or CNN models. Additionally, the project will explore the integration of these transformer models with reduced-order models, like autoencoders, to manage the complexity of high-dimensional systems effectively.
Experience required: self-motivation, passion of research, good coding skills with pytorch (in preference) or tensorflow.
Contact details: Dr Sibo Cheng, Research Associate, Dept of Computing, Faculty of Engineering, South Kensington campus. Email: email@example.com