Students in the first cohort of the UKRI Centre for Doctoral Training (CDT) in Safe and Trusted AI (STAI) have been working with our partner, Royal Mail, to investigate how AI techniques and telemetry data could be used to improve the trustworthiness of automatic route change suggestions resulting from motorway closures.
The first group project was undertaken in partnership with Royal Mail with some of their staff coming to the CDT’s university site at King’s College London to discuss and explain details of the problem, and CDT students visiting Royal Mail’s CPC Delivery Offices in Camden to discuss their work. Commenting on the project, Dr Alexander Green, Senior Data Scientist for Royal Mail’s Group Business Intelligence, said the students had “a whole range of insights and original research we wouldn’t have had time to complete ourselves!”
“The freedom that the CDT students have to explore different avenues of thinking that either didn’t occur to us or that we hadn’t explored fully yet was a great way for the students to add value to the partnership with Royal Mail,” he added.
Engagement with a broad range of non-academic partners is a key component of the CDT. Dr Chris Hampson of King’s College London, who coordinated the group project said: “The group project was a great opportunity for our students to get hands-on experience in seeing how principles of safe and trusted AI can be applied in real-life industry settings, and in proposing suggestions to improve the safety and trustworthiness of intelligent systems that are vital, and often invisible, in everyday life. Working with Royal Mail is one way in which we provide assurance that both the research undertaken by the CDT and the skills developed by our students will be relevant and valuable to industry and society at large, while also informing and supporting UK industry in producing state-of-the-art safe and trusted AI solutions.”
After restrictions resulting from the COVID-19 situation were imposed, all activities involved in the group project work were successfully transitioned to remote working, and the students completed their projects virtually, culminating in the presentation of their findings and suggestions for potential improvements to our Royal Mail partners and CDT staff and students. Jazon Szabo, a first year CDT student in the 2019 cohort, said, “What I personally liked the most about the group project was that it exposed us to how our industry partners work and what kind of problems they spend their resources solving. More importantly, it showed us how we could apply the skills that we have in an industry setting, which I find very important and useful for our future careers.”
The Group Project
As part of their training activities, first year students from the STAI CDT are required to undertake a group project with one of our industrial partners. The aim of the group project is to identify how cutting-edge research in model-based AI techniques could help to make some aspect of one of our partner’s operations safer (in the sense of having some assurances over its behaviour) and/or more trustworthy. The experience of the group project supports students to develop a range of skills, including: teamwork; time management and planning; product management and working with stakeholders; report writing; presentation skills; analytical skills; requirements gathering skills; and an understanding of relevant state of the art AI techniques.