What are the opportunities to derive insights from cellular network data?
Introduction
The ongoing evolution of telecommunications technology has created a wealth of data that could revolutionize decision-making across various sectors. Mobile phones, with their ubiquity and constant connectivity, generate vast amounts of valuable information that, if harnessed effectively, can significantly enhance planning and decision-making.
This data and its subsequent analysis could replace or complement traditional (movement) survey methods for estimating trip volumes and destinations, commonly known as Origin-Destination (OD) trip matrices.
This GJ Consulting discussion provides an introductory exploration of the many opportunities presented by cellular network data in the fields of transport planning, urban planning, city planning, and event planning.
We will explore other specific applications for telecommunications data in future GJC articles.
Background - cellular (mobile) phone movement data
Cellular network data has emerged as a powerful means to support traffic planning and decision-making in transportation network development. This data source can provide a detailed understanding of travel patterns, easily surpassing the limitations of traditional travel surveys. Additionally, its optional real-time updates could offer a comprehensive operational view of mobility patterns across various travel modes.
Making use of cellular network data for traffic analysis requires identifying movements within raw data to detect meaningful trips between different locations. Algorithms can be used to infer travel demand from these (extracted) trips, as well as other algorithms for route estimation. It is, however, usually necessary to validate and evaluate algorithmic performance.
Research to date has highlighted weakness around the application of algorithms to existing cellular network data and the impact of low resolution and 'high noise levels' on results. The choice of trip extraction methods hence influences the outcome, with short walking and cycling trips proving more challenging to capture accurately, and potentially leading to under-representation.
Privacy concerns are frequently cited as an real barrier to the use of cellular network data - but active use cases are showing increasing confidence in efforts to protect the privacy of individuals (while still extracting the utility of the data). This includes through the use of pseudonymization (taking identifiable data and replacing it with a value that cannot be linked without additional information located elsewhere), and anonymization (replacing original data with a value both unrelatable to the original data and irretrievable).
Despite the existing limitations with data resolution and noted privacy considerations, it is clear there is significant potential for cellular network data to provide a detailed picture of aggregated travel patterns across numerous (transport) modes. The ability to gain insights into individual origin-destination pairs, including time profiles for specific zones, almost certainly surpasses the capabilities of traditional travel surveys.
While research demonstrates the significant potential of cellular network data, further exploratory work is ongoing. Investigating travel patterns by different modes, activities, and demographics remains an active area of research, requiring additional work to ensure the representativeness of results for wider population groups. The absence of large-scale 'ground truth data' also remains a challenge in many cases, as this limits comparative metrics to validate findings. This will no doubt improve over time as base data is established.
Further opportunities to trial could include real-time applications using cellular network data. This could support active operational work in a wide variety of ways. Cellular network data could then be used as the primary source which is then layered with other data sources such as sensors, point of sale data, census data, GPS tracks, or travel surveys. In saying that, it is clear that further research needs to be undertaken around the integration of travel patterns from cellular network data and existing traffic models.
Opportunities
Transport Planning. Mobile cellular data could play a pivotal role in optimizing transportation systems. By analyzing mobile phone signals, transport planners could gain insights into traffic patterns, identify congestion hotspots, and optimize routes in real-time. This could enable more efficient transportation planning and resource allocation.
Emergent companies like Equity AI (Equity AI Ltd) highlight the possibilities for transport planning. Equity AI uses data, artificial intelligence, and movement analysis to provide near real-time insights into traffic movement and modes to allow for improved modelling, scenario testing, and better decision making. Equity AI utilises cell phone and other data in a specific way to derive insights.
Urban Planning: In urban planning, cellular data could be utilized to understand population movements, peak activity times, and popular areas. This information could enable the design of urban spaces that cater to the needs and preferences of residents, fostering sustainable and well-connected communities.
Challenges
Despite the immense potential, the use of telecommunications data for decision-making poses risks and challenges, most notably in the realm of privacy. Adhering to regulations such as the EU GDPR is crucial to protecting citizens' privacy rights. Striking a balance between leveraging data for societal benefit and respecting privacy is a challenge that demands careful consideration.
Data accuracy is also acknowledged as an area needing attention with mobile phone tracking data, particularly for short-distance trips. Caution interpreting data and further research in this area is hence needed along with comparative data to validate analysis.
A final challenge has been the reality that the opportunities from mobile phone signalling event data is frequently overstated and then under-delivered. The reason is twofold:
operators don’t normally have inhouse experience on mobile telecommunications radio protocols and so they don’t know what data to capture. Then, if if they do have that knowledge, they often don't have the ability to decoding and hence cannot accurately detect trips (including short trips).
even when the operator has the radio protocol events, geolocating them is a complex undertaking that operators frequently rely on vendors to do.
Conclusion
In conclusion, the opportunities presented by telecommunications data for decision-making are huge and potentially transformative. There are a range of compelling use cases in a number of sectors including transport planning.
That said, these opportunities must be approached with a firm commitment to privacy and ethical considerations. Piloting emerging technologies is essential to test specific use cases and refine methodologies, address challenges, and build trust with data holders and the public.
In the emergence of 5G networks, edge computing, AI, and other disruptive technologies, telecommunications providers have a range of new income-generating opportunities. Monetising cellular network data seems an obvious and inevitable path as long as it is undertaken in a measured and safe manner.
To harness the full potential of mobile cellular data, it is recommended telecommunications companies and network operators find trustworthy partners to develop these niche areas and initiate pilot programs. This will allow for the iterative development of good practices, ensuring that the significant opportunities from cellular data does support strategic and operational planning applications.
References
Graham, S., & Marvin, S. (1999). Planning cybercities? Integrating telecommunications into urban planning. The Town Planning Review, 89-114.
Caceres, N.; Wideberg, J.; Benitez, F. (2007). Deriving origin-destination data from a mobile phone network. IET Intelligent Transport Systems, 1(1):15 – 26.
Calabrese, F.; Di Lorenzo, G.; Liu, L.; Ratti, C. (2011). Estimating origin-destination flows using mobile phone location data. IEEE Pervasive Computing, 10(4):36.
Calabrese, F.; Diao, M.; Lorenzo, G. D.; Jr., J. F.; Ratti, C. (2013). Understanding individual mobility patterns from urban sensing data: A mobile phone trace example. Transportation Research Part C: Emerging Technologies, 26:301 – 313. doi:10.1016/j.trc.2012.09.009.
Comments