Project Summary

PolyChord, an advanced Data Analytics company, collaborates with Focus Sensors in the TRIG project to analyse data from line-side fiber optics and passengers’ smartphones (with permission) to identify the causes of poor passenger experience and indicate pinpointed fault spots on the track for inspection. Transport operators highly anticipate the development of this cost-effective tool to direct teams to fix identified problems.

Project Achievements

The project focussed on collecting mobile phone and FOMS data from the same route and time. With data, we could then proceed with a feasibility study to establish how we could use FOMS data to get a better fix on the location of issues detected by mobile phones with PolyChord’s RoRi tool. We successfully collected the necessary data and ran the RoRi pipeline on the phone data which allowed us to establish methods that connect a RoRi outlier detection with the position of the train, and therefore identify the exact location of the issue by focussing analysis on the location and time of the train. We also established methods to synchronise the data even when the timestamps are not synchronised. The most important outcome of the project in many ways is the learning that we need to design our mobile-phone data collection app so that it records the GPS satellite time rather than the phones local time, to enable easy synchronisation of datasets. The other important outcome was the drastic boost to PolyChord’s relationship with LNER and Network Rail, resulting in a grant application and now discussion about East Coast partnership funding.

Conclusions

We have concluded that outliers from RoRi can be associated with train position using timestamps, and then the location of a fault more accurately pinpointed by extracting the section of FOMS data associated with a particular train position and performing outlier and/or correlation analysis on this data. This will inform the future development of RoRi and will ensure that we record the satellite time rather than phone time in recording the phone sensor data. We have also developed a framework for dealing with instances where the FOMS unsynchronise with the satellite for any reason. The project is considered successful.

Next Steps

We are in discussions with the East Coast Partnership to fund the next phase of development of RoRi, which is to develop a tool that uses customer mobile-phone data and develop the API needed to collect this data via the LNER phone app. In the meantime, we will apply for the next Smart Grant a nd any other funding we identify that is suitable. We will also look to other countries with rapidly expanding railway networks, who may have more money and can be quicker moving, such as the UAE, to develop the system. TRIG has provided us with invaluable learning that will inform future development of the RoRi tool and has drastically improved our relations with LNER, Network Rail, and our TRIG collaborators Focus Sensors.