Coventry University Featured PhD Programmes
John Innes Centre Featured PhD Programmes
University of Kent Featured PhD Programmes
University of Liverpool Featured PhD Programmes
Bournemouth University Featured PhD Programmes

Robot homing deeply reinforced by another robot

This project is no longer listed on and may not be available.

Click here to search for PhD studentship opportunities
  • Full or part time
    Dr A Altahhan
  • Application Deadline
    Applications accepted all year round
  • Self-Funded PhD Students Only
    Self-Funded PhD Students Only

About This PhD Project

Project Description

The Project

The motivation for this PhD proposal is the research performed by Dr Abdulrahman Altahhan for more than 10 years on robot homing.

Visual robot homing using vision is one of the important problems in robot navigation and has been recently under extensive focus. Yet to learn to navigate towards a goal using agents’ visual capabilities where the goal is not on sight and without intervention or environment engineering is difficult. The problem pertains to all animals naturally and is a must for most of the commercial and entertaining robotics application. Central to this ability is the skill of orienting towards the home when it is off sight and recognising it once the agent is around it. Animals do that by wiring the surrounding visual memory somehow to their neural map of the environment. How they do that is yet to be discovered. Traditionally it has been linked to distinctive position or places in the environment i.e. landmarks. However, the way animals do its navigation and find their home suggests something more subtle than only landmarks.

In this project we will be investigating new ways to do homing using artificial neural networks and reinforcement learning. In particular we will be implementing a model that should allow a robot to be deeply guided by its internal reward mechanism as well as other external reward sources, such as another robot, on how to reach a goal. Recent development in reinforcement learning shows that this is more possible than ever. Deep learning will be used to extract interesting features from the robot sensors which is assumed to come in high velocity. A deep reinforcement learning architecture is going to be used to sustain the relationship between the learner and the guider. The new architecture is envisaged to be similar to the actor-critic architecture with superior properties for more practical learning scenarios.

Duration: Full Time 3 years Fixed Term or Part Time 5 years Fixed Term

About the Centre/Department

The computing department at Coventry University offers world-leaving research in computer science and informatics. Our strengths are in the fields of Distributed Systems, Computational Intelligence, Serious Games and Cybersecurity. Members of national and international teams, our professors, readers and senior academics work with industry to reach new heights of discovery in computer science creating innovations that benefit society and solve global problems. We are seeking top quality graduates to help us in this quest. Excellent supervision in a nurturing, vibrant research environment will be offered to successful candidates in order that they can reach their highest potential when pursuing the PhD programme.

Successful Applicants

Successful applicants will have:
- A minimum of a 2:1 first degree in a relevant discipline/subject area with a minimum 60% mark in the Project element or equivalent with a minimum 60% overall module average, or

- A Masters Degree in a relevant subject area will be considered as an equivalent. The Masters must have been attained with overall marks at merit level (60%). In addition, the dissertation or equivalent element in the Masters must also have been attained with a mark at merit level (60%).

- The potential to engage in innovative research and to complete the PhD within a prescribed period of study.

- Language proficiency (IELTS overall minimum score of 7.0 with a minimum of 6.5 in each component).

Find out how to apply:

See the website:

FindAPhD. Copyright 2005-2019
All rights reserved.