Jnction Awarded Innovate-UK BridgeAI Contract

Jnction has been awarded a contract to carry out a feasibility study into the use of AI techniques to improve predictions of train service delays for the UK railway. The study will be completed by 31st March 2023. The work will be carried out in partnership with InterimConsult’s Paul Corcoran, who has over 20 years’ experience in AI and the rail domain. 

Prediction of train delays is a major issue for the UK Rail Network. According to the Office of Rail and Road, between October and December 2022, delay minutes attributed to Network Rail increased by 18.7% compared with the same quarter the previous year, and only 62.3% of trains ran on time. 

An improved delay prediction methodology using AI could therefore have a significant impact of the rail industries punctuality and both improve customer experience and improve performance, when applied to train operations. 

Train delays often result from unforeseen events such as infrastructure issues, weather conditions, and operational disruptions. By developing an AI-based system that can accurately predict potential delays in real-time, operators can proactively respond, minimise the impact on service quality, and improve accuracy of passenger information and hence customer experience. 

This study will assess the technical, economic and commercial viability of the proposed solution, laying the groundwork for a future transformative application within the rail industry. If successfully implemented, Jnction’s Bridge AI could enhance overall performance of train services, minimise disruptions and improve customer satisfaction.   

Study Components:  

Jnction’s feasibility study will employ a combination of data analysis, machine learning techniques and mathematical modelling to develop a predictive model. The UK rail system is currently run by old legacy IT systems and needs a significant modernised upgrade. 

1. Data Analysis: Comprehensive data sets will be collected and analysed to identify the key variables which influence train delays, such as weather conditions, maintenance schedules and historical performance.  

1. Machine Learning Model Development: Advanced AI algorithms will be developed to predict train delays based on the identified variables. 

1. Performance Evaluation: Jnction’s AI model will be evaluated using simulated scenarios based on historical train data. This will be assessed to measure the effectiveness of the solution. 

Why Now?  

AI was identified as one of the 5 critical technologies in the 2023 UK Science and Technology Framework by the UK Government. This publicised support for further development into AI highlights the government’s desire for the UK to become a world-leader in innovation and technological expansion. This was backed in March 2023 by Jeremey Hunt, the Chancellor of the Exchequer, who stated that “the government [has] ambitions to make the UK a scientific and technological superpower.” Hunt set £3.5 billion aside to support projects including AI-based systems to help grow the UK economy.