My interest in user-centred design began when I was working in Digital Marketing and Communications. Since venturing into Human Factors I have become interested in sociotechnical systems in infrastructure and applying inclusive design principles to safety critical products.
If you have spent some time in UK National Rail stations you will be familiar with an automated voice telling you where your train will depart from, to report anything suspicious, and stand behind the yellow line. Spoken announcements (SAs) are a tool UK railway stations have been using for decades to make the station more accessible for sight-impaired passengers and to manage passenger safety and security. Yet, there is little research on how effective this tool is in making passengers behave safely. Most passenger accidents and fatalities in 2024/25 happened in a station; the passenger fatality risk at the platform-train interface was 52% and over 3,300 slips, trips and falls were reported to the Rail Safety and Standards Board (RSSB, 2025).
The aim of the research was to understand passengers’ and industry experts’ attitudes to safety and security spoken announcements (SAs) as a means of identifying features that make them effective. Three research questions were generated to meet this aim:
A mix of qualitative methods was chosen to explore passenger and expert attitudes toward, and experiences with, announcements. There were 53 participants in this study.
Passengers (n =37) at two mainline rail stations, London Paddington and Nottingham, were surveyed. Most questions were open-ended to allow passengers to elaborate on their opinions and experiences. 65% of participants were infrequent passengers, 22% were frequent or very frequent passengers, 13% were causal passengers.
Semi-structured interviews were conducted with staff (n = 9) at London Paddington, Nottingham, and Bath Spa stations. Participants were asked about the use of manual SAs at their station, how different passenger groups respond to SAs, the tools they use to manage crowding and protocols for managing emergency evacuations, and what they would change about current SAs. A focus group was conducted with members of the RSSB People on Trains and in Stations Risk Group and Confidential Incident Reporting & Analysis System (n = 7). 50% of participants in the expert group had experience making announcements. The median years of industry experience was 12.75 (range: 3 – 25 years). Thematic analysis of all the data from transcribed interviews, focus group and survey responses was performed using NVivo 14.
Covert station-based observations were conducted at Paddington and Nottingham. The type, mode, frequency of SAs and behaviour following SAs were recorded as well as crowd density and flows using a modified R.A.M.P analysis.
The themes that emerged from the passenger survey were: relevance, inspires action, personal safety, personal security. The themes that emerged from expert opinions on effective SAs were: staff authority (which was dominant), style of delivery, tailored address.
Announcements are one tool within the socio-technical system of a rail station influenced by a broader rail network that is challenged with meeting regulation requirements to remind passengers to be vigilant of danger while reassuring them they are safe so they continue travelling by rail to meet financial goals. The station architecture affecting sound propagation, environmental noise, especially from trains at standstill, and poor speech intelligibility are barriers to engagement with SAs. Station staff having minimal control of automated announcements means there is the potential for conflicting messages from manual and automated announcements during service changes which can lose passenger trust.
Passengers find security SAs more useful than safety SAs, though most do not recognise they are also responsible for station security. Passengers and experts agree specific, timely announcements get more attention, which is why manual SAs are preferable to automated safety and security SAs that can fail to be contextually congruent and cause habituation as a result.
I conducted a systems analysis of the causes of the 2017 Grenfell Tower Fire. Using an accimap and actors map, I analysed the findings of the official inquiry and found the main causes of the tragedy were:
I assessed the physical ergonomics of three tasks by a bread baker: loading large sacks of flour onto shelves, kneading dough and moving goods in and out of the oven, and assessed the risk of developing musculoskeletal disorders. I conducted postural (REBA and RULA), anatomical, biomechanical, physiological and anthropometric analyses. The primary risk to the bread baker was developing lower back pain from standing and bending, and wrist injuries from repetitive motions. I presented the following recommendations:
Along with colleagues, I designed a medium fidelity prototype of a mobile application for people who have not adopted mobile banking or investing or e-wallets and find the benefits of using them unclear. The app had a simulated investment section to help users practice investing. I did desk research on the target demographic, created user personas, conducted in-person, moderated usability tests with participants fitting or closely fitting the intended user profile and redesigned elements of the app users struggled to use by improving clickability signifiers and information architecture.
