New Edge Computing Solutions for Machine Learning

  • New Edge Computing Solutions for Machine Learning

School of Electronics, Electrical Engineering and Computer Science
& ECIT Global Research Institute

Proposed Project Title: New Edge Computing Solutions for Machine Learning

Principal Supervisor:   Professor Roger Woods         Second Supervisor:    Dr Anna Jurek

Project Description:

A key research effort has been expended in employing artificial intelligence (AI) approaches to solve the major problems facing mankind. Many of the world’s major companies have been developing technology to achieve this. For example, Google’s DeepMind and Facebook’s AI Research (FAIR) have been applied to applications such as mental disease diagnosis, improving visual and audio recognition, reducing patient diagnosis to treatment times amongst others. In many cases, the process involves capturing a wide range of data (information) into a centralised AI resource which acts to learn behaviour thereby allowing prediction of future activity and detection of anomalous behaviour.

Edge computing will be the next major step in AI computing as it will act to overcome communications delays. Traditional cloud services provide limited functionality for the growing number of information devices, particularly as they cannot always handle the latency involved with sending data back and forth between the end user and the cloud provider's data centre. One alternative is to start putting more processing power onto devices at the edge of the network to serve customers sensitive to those delays. For example, Intel have introduced the Movidius Neural Compute Stick which aims to deliver processing power optimised for artificial intelligence into edge devices like cameras. Hence, we are starting to see the shift towards AI-enabled edge computing technology particularly, in distributed and remote environments.

This PhD project will look to either exploit such technology or develop embedded FPGA technology to develop new forms of AI solutions. This will be exploited for applications highlighted by companies such as Analytics Engines Ltd., Adoreboard or Bazaarvoice with whom the team are working closely. The project will tackle the engineering practical problems associated with edge computing complexity and delay associated with this approach. The overall goal will be to improve the performance of some practical applications. The successful applicant will have access has cutting-edge facilities and will develop highly valuable skills in AI and edge computing solutions.

Contact details

Supervisor Name:               Professor Roger Woods                                             Tel:         +44 (0)28 9097 4081
QUB Address:                     Ashby Building, Stranmillis Road, Room 07 021          Email: 

2nd Supervisor Name:        Dr A Jurek                                                               Tel:          +44 (0)28 9097 5488
QUB Address:                    Computer Science Building, Room 3.009                     Email: