A Real-time Refresh on 2011 Census Stats using Huq’s Geo-data

By Ellie Marfleet, University of Leeds final year BA Geography student.

The problem with using the most recent census data to study regional commuting patterns is that it is almost a decade old. The potential of using new big data sources are evident, as companies like Huq record movements of people based on their differing locations throughout the day. Similarly, the promise of this data is more efficient than conducting traditional surveys which, it could be argued, are out-dated as soon as they are printed.

During my final year at University of Leeds I have been able to delve into different data sources used to analyse commuting patterns in West Yorkshire as part of my undergraduate dissertation. The research process that I have pursued differs from many on my degree programme as I chose to partner with a private sector company. I began my preliminary research by exploring the geographical issues and topics likely to pervade the new decade as the focus of my project.

After reading about the trial conducted by the Office for National Statistics (ONS) regarding the use of mobile phone data (MPD) to estimate commuting flows, it became obvious to me that the use of these administrative datasets would be crucial in understanding human populations and trends over the coming decade. The ONS used MPD collected by extracting the locations of masts/cell-towers frequently connected to by mobile devices, which operators then used to estimate the geographical areas containing the usual residence and place of work of users.

My research process advanced after a discussion with Steve Halsall of Red Tiger Consulting, regarding the availability of Huq’s geo-data data for research projects through their CMO. I then explored the methodology behind Huq’s dataset resources, and the benefits of using Huq’s data became clear. Within their dataset, each geo-data event comprises location co-ordinates and a timestamp among many wider attributes, and is updated daily (huq.io/products/#build-models-features). This is in contrast to the more inference-led insights produced during the ONS study, which is likely to generate less accurate results.

Since Steve’s referral to Huq, I have been thoroughly supported through both the refinement of my project goals and production of the specific dataset necessary, which is bounded to 2019 (versus 2011!) and spans eight specific workplace zones as destination outputs. Huq’s enthusiastic support throughout the research process has been incredibly helpful, in addition to the speed at which they provided the data following my request. I look forward conducting a presentation later this term to second year students on my positive experiences with Huq, in hope of further reinforcing the value of private sector collaboration and support potential through and around their courses at the University of Leeds.

Ranging from commuting and journey times analyses to trends in footfall and even road congestion, the versatile applications of Huq’s data should appeal to students, researchers and academics from a host of undergraduate and postgraduate degree disciplines. I look forward to progressing with my research project using Huq’s geo-data and to sharing the outcomes of that in due course.

A massive ‘thank you’ to the team @ Huq and to Steve Halsall @ RedTiger for their support throughout my research process and hope that others can gain similarly from this experience like I have.