Congratulations on considering getting yourself a PhD! The fields of artificial intelligence and robotics are progressing more rapidly now than ever before, and this is a very exciting time to be carrying out state-of-the-art research at a world-leading university. Graduating with a PhD from my lab will open up a range of unique opportunities, including highly-competitive jobs across the top tech firms, academic positions with a view towards one day setting up your own research group, and startup ideas for those with entrepreneurial ambitions. As a PhD student in the Robot Learning Lab, you will be conducting exciting and ambitious research in a supportive and friendly environment, with state-of-the-art facilities and regular interaction with me and other researchers, and you will have the opportunity to travel internationally to the top conferences and events across the world.
How to apply
Every year I offer a small number of PhD positions in my lab, for those who wish to explore how computer vision and machine learning can be used to train robots to perform real-world object manipulation skills. Successful applicants will typically have a Master's degree from a high-ranking university, with a grade equivalent to a UK distinction, and will be able to demonstrate practical experience of robotics, computer vision, or machine learning. Candidates should formally apply through the Department of Computing's online application system, and guidelines for this process can be found by clicking here. In the "Academic Programme" field, you should select "Computing Research PhD" (do not select "AI and Machine Learning PhD 4YFT"; this is for the AI4Health CDT ). In the "Proposed research supervisor" field, you should enter "Edward Johns", and in the "Proposed research group" field, you should enter "Robot Learning Lab". In the "Proposed research topic field", you should enter a title of your choice, based on your research proposal.
In most cases, new PhD students start in my lab at the beginning of October, and I am currently accepting applications for October 2020 intake. Formally, there are three application deadlines: 1st November 2019, 10th January 2020, and 13th March 2020. You may make your application at any time of the year, but you will not be notified of any progress until after the deadline by which you apply. Once the deadline passes, applications will be assessed by our admissions committee, and promising candidates will then be invited to an interview. If we decide to make you an offer after your interview, then this will be conditional on you securing funding for your PhD. Some students secure their own funding, but most require a funded position, and in this case your application would then be passed to the department's funding committee for further assessment.
For students seeking a funded position, this would cover tuition fees and living expenses in London. There are a number of different funding sources, which depend on your nationality. For British or European citizens, the Department of Computing has a small number of funded scholarships every year (European citizens are still currently eligible after Brexit). For non-European citizens, there are a number of country-specific scholarships, such as the China Scholarship Council. Further information on scholarships for international students can be found by clicking here. Some international students may have already applied for these scholarships themselves, and if you have already received a funded scholarship, then you should specify this on your application form. Otherwise, you should specify on your application form which scholarships you are eligible for, and the department's admissions panel will then decide which scholarships to nominate you for. Following this, funded scholarships will then be offered to the highest ranked applicants.
Your application should include a research proposal, detailing your thoughts on a new way in which computer vision or machine learning can be used for robot manipulation. This will be used as part of our assessment of your application, and will also form the basis of discussions should you be invited for an interview. If you were to receive an offer, your actual PhD may vary from this proposal, based on my guidance and recent developments. So, rather than being a precise plan for your PhD, the proposal is your chance to showcase your ability to be innovative and rationalise your ideas. Your proposal will be assessed primarily by me, so please write it with a specialist audience in mind. You are free to decide the length and format of your proposal, but I recommend being concise and writing no more than four pages.
Examples of current topical areas of research are: (i) data-efficient model-based reinforcement learning for real-world robot manipulation, (ii) transferring control policies trained in simulation over to the real world, (iii) learning of new tasks from a small number of human demonstrations, (iv) visual state estimation for object grasping and interaction, and (v) visual scene understanding for planning multi-stage tasks. For further ideas, you may also wish to read up on published work in the lab by clicking here.
Please note that we focus primarily on robot manipulation and interaction from visual observations, and successful candidates will be those whose proposal addresses real-world learning problems, such as when dealing with high-dimensional image observations, and the experimental challenges of reinforcement learning with physical robots. The best proposals I receive are those which combine learning-based methods (such as reinforcement learning) with classical methods (such as optimal control and vision-based state estimation).
Try to be creative in your proposal: rather than simply describing background theory and existing methods that have already been published, I encourage you to be brave and describe a new idea you have been thinking about, even if it is preliminary or speculative. Of course, you are not expected to be an expert in the field yet, but you should show that you have sufficient motivation to read around the subject and learn about the state-of-the-art, and sufficient creativity to propose novel ideas which address limitations of existing methods.
Before making your application, please do get in touch with me so that we can discuss whether your interests are aligned well with those of my lab. For this, or to ask any informal questions about the application process, or just to find out a little more about life in my lab, then please email me at firstname.lastname@example.org. So that I know you have read this webpage, please use "PhDApp2020 Xyz Xyz" for the email's subject line, where "Xyz Xyz" is your name.
Thank you, and I look forward to hearing from you!