We have moved the mystery of data science consulting into reality for Network Rail and enabled their staff to be confident in using data to improve the journeys of millions of people per year.
Network Rail South East (NRSE) had minimal understanding of sub-threshold delays (less than 3 minutes) due to the existing monitoring method capturing delays in minutes instead of seconds, and subsequently not conducting a detailed root cause analysis. However, in 2016, 40% of delay minutes were sub-threshold delays. With 500 million passenger journeys per year serving the economic hub of the South East, small delays impact every part of life. NRSE therefore required a better understanding and visibility of small minute delays to determine the most effective places to implement measures.
Amey Consulting set out to investigate under-used existing data sources within NR and turn this data into actionable insights, enabling network staff to understand the importance of "seconds performance". To bring about the most change within the project Amey took a high velocity approach, managing projects using agile techniques, ensuring that approaches and methodologies were subject to frequent formal reviews. This allowed for the implementation of immediate changes, maintaining a solid pace for continuous improvement.
Our consultants brought data alive by working in a way that an operator could understand and own – we demystified data and as practical consults, we helped upskill a new breed of operational data staff. We built a cloud-based data processing and storage solution, processing internal data in real-time, and displaying as information on an interactive platform. Our team of skilled mathematicians turned data analytics into insights for Network Rail. The final result of this engagement is called 'Quartz': a tool which utilizes to-the-second train movement data to present the performance of individual stations across the South East route, allowing users to understand the impact of small process changes on performance.
The outcome of the project exceeded the expectations of Network Rail and showed them the possibilities of significantly improving the performance and operations of trains. 50% of Network Rail users are now using Quartz for analysis at least 1 to 2 days a week. An example of the successful implementation of Quartz can be seen in the changes taken place at East Croydon station. Quartz enabled Network Rail to identify that platforms 3 & 4 were the main causes of delay. Upon further investigation it was found that this was a result of staff inability to cope with multi-dispatch during peak hours. As a result 6 additional full-time staff were added, resulting in a 40% decrease in daily delay minutes, saving £3.5million over the next 5 years.