A Case-Based Real-time Adaptive Engineer Site Support System
Robert Gordon University & British Telecom
Closing Date: 10.00am 1 August 2016
Applications are sought for a Research Studentship (PhD) in Machine Learning and decision support systems. The project will be up to 36 months duration, commencing in October 2016. The studentship is funded by British Telecom (BT) and includes Home/EU tuition fees as well as a tax-free stipend of £14,057 per annum for three years. Non-EU students may also apply but on a fees-only basis.
The aim of the PhD project is to provide real-time assistance to field engineers while they are executing a task on site jobs using knowledge management strategies driven by Machine Learning and Case-Based Reasoning (CBR). Information, knowledge and experiential content will be captured and maintained in a casebase repository to create a key corporate asset to be harnessed for knowledge management, access and sharing.
Problem solving in complex business scenarios increasingly requires large amounts of specialised knowledge. Typically such knowledge in the form of experiential content remains in the heads of employees and so are not easily accessed, shared or distributed. The CBR methodology provides the processes to capture, store, retrieve and reuse such experiential knowledge for future problem solving. Such a resource can then be utilised to recommend engineers (based on relevant experience and skill levels) to BT service delivery jobs as and when they occur and provide real-time assistance on the job. This creates an intelligent real-time support framework for BT’s engineering workforce.
The initial phase of the PhD will involve familiarisation, literature review, and a pilot mini-project to collect experiential data from BT to create a CBR demonstrator. RGU’s existing experience with CBR development software will be shared with this PhD project. There will also be opportunities for the student to have access to BT user groups (engineers) associated with this project. Subsequently the student will focus on specific structured and unstructured content (including text, image and video) with a view to identifying the best representation to create cases. It may well be that state-of-the-art work in feature engineering from deep learning research and the use of virtual reality for experience sharing will also be explored.
The student will initially spend a considerable amount of time in Ipswich with BT researchers, benefiting from direct access to BT engineers and data sources for data collection and also participate in research group meetings as appropriate. We expect that from the second year onwards the student will benefit from daily access to expertise from postdocs and other PhD students at RGU, weekly meetings with supervisor, activities such as the fortnightly CBR group meetings, research away-days, annual PhD symposium, and SICSA events. Skype with both supervisors will initially be fortnightly and thereafter be monthly, and joint quarterly progress review meetings will alternate between RGU and Ipswich.
Applicants should have (or soon expect to have) a good Honours degree in Computing (1st or 2.1) or a Masters in Computing at Distinction level. Strong programming skills are highly desirable. Some knowledge of AI, in particular Machine Learning techniques such as supervised and unsupervised learning models and / or Case-based reasoning methodology. Interest s in deep learning and virtual reality is also highly desirable, though not essential. Applicants should have good personal and communication skills, strong professionalism and integrity and be confident working on their own initiative.
Applicants should Apply Online by 10.00am on Monday 1st August. When applying, click on advertised studentships and select CSDM-16-1. Please note that the start date is OCTOBER 2016 (even though the apply online system may automatically indicate that it is too late to apply for Oct 2016 entry). Please submit a personal statement of interest in place of a research proposal. Interviews will take place in the week commencing 8th August 2016.
All enquiries should be addressed to Dr Nirmalie Wiratunga, email@example.com
One can download a pdf of the above details (here).