Introduction
From Luisa Moisio, Director of Research and Development
Of the projects featured in this edition of the Research Highlights, it’s striking how many of the findings and solutions depend on data accessibility and making use of new ways to generate insights from data.
The new PREPAIR delay prediction tool currently being trialled in the National Rail Communication Centre is a machine learning model. It was trained on historical delays, which it applies to live Network Rail data in order to generate predictions on how delays develop. On the subject of safety-critical communication, a successful proof of concept has raised hopes that using artificial intelligence (AI) could lead to fairer, more thorough assessments of how frontline staff communicate, thereby enabling the detection of issues and improvements in the focus of interventions and training. And one of the biggest successes of our strategic partnership with the University of Huddersfield, now celebrating its first ten years, is Red Aspect Approaches to Signals (RAATS). This provides vital insights on signals approached at red to the industry by combining many datasets with advanced algorithms.
Challenges remain in opening up the vast reservoirs of data from the everyday running of the railway and in generating actionable insights from the data available. From the start of 2022, with the establishment of the Data, Systems and Telematics Standards Committee, there is a growing recognition of the essential role of common data formats and standards in sharing data more easily and exploiting it more effectively.
As AI becomes an ever-growing influence in almost all areas of life, we need guidance to help industry make the most of the potential for different AI techniques in building a safer, more sustainable, and customer-focused railway. One of the many risks associated with the growing use of AI is that humans ‘zone out’ because their active interventions and higher cognitive skills are more rarely needed. Still, when a failure or an emergency does occur, humans need to be ready to act and perform at their best. Our work on managing cognitive underload in train drivers is an early example of the types of challenges that a different balance of human–machine interaction will bring.
Having agreed a high-level research framework for Control Period 7, in which digitalisation and data feature strongly, we are working together with the Network Rail R&D team on the detailed needs and opportunities that we will collectively focus on. So, this autumn is the right time to get in touch and share your ideas for future research. I look forward to hearing from you.
Luisa Moisio, Director of Research and DevelopmentLuisa.Moisio@rssb.co.uk
RSSB’s cross-industry Research Programme enables a safe and efficient railway by providing knowledge and solutions to best utilise existing assets and unlock future change.
The research challenges and opportunities we focus on require strong collaboration across the rail industry and benefit greatly from the steer and coordination that RSSB provides, with its independence, strong technical capabilities, and world-class risk modelling and analysis expertise.
The Research Programme is predominately funded by the Department for Transport. We also owe a big thanks to many organisations across the industry that provide significant in-kind support to research, co-fund work with us, and pioneer translating the findings into positive change on the railway.
Much of our research supports the Rail Technical Strategy, which comprises five key priorities. We also research topics to help improve the health and wellbeing of railway staff and those who use the network.
Where our research specifically feeds into any of those, we use these icons.
