Innovate UK: NEXUS - Real Time data fusion and Network Analysis for Urban Systems

Lead Research Organisation: University of Oxford
Department Name: Oxford Internet Institute

Abstract

This project will combine multiple data sources based on Oxford City and its social and economic development requirements. The team comprises BT as lead partner, to provide advanced data analytics and visual software tools, plus access to a number of networked large-scale data sources. The academic research partner is the world-leading Oxford Internet Institute (OII), who will lead on novel graph and network visualisation analysis. The end user is Oxford City Council, who have expressed a keen interest in the project outputs; as the city faces a number of challenging constraints in the coming decade. Driven by economic growth, population increase, housing demand, and serious transport issues. The city must also balance intense demands from tourism, and social provision for its large student population. The project will aim to integrate a number of social, economic and transport related data streams in real-time to provide the city with a revolutionary geo-spatial viewpoint and perspective into its emerging challenges. This new capability will facilitate city planning, resource management, social-economic development, and reduced pollution/congestion from improved transport operations

Planned Impact

Not required as per NERC instructions
 
Description There were two main things developed through this grant. First, we developed novel means of estimating local commuting patterns from social media data. Accurate information about transport patterns is vital for effective local development of transport infrastructure. Yet we actually know very little about who is travelling where (especially on road networks). This makes it hard to take effective decisions (for example, about which bus routes should be subsidised, or where to prioritise new road networks). Our work looks at whether data freely available from the social media platform Twitter could be used to generate estimates of local commuting patterns. We developed a simple model using geolocated Twitter data which we demonstrated could perform better than existing standard models for estimating commuting flows. Second, work focussed on novel methods for visualising big, open and social data in a way that makes it easy to use for policymakers and practitioners. We developed a novel algorithm for producing "tile maps" (maps where each geographical unit have the same size - for example each country on earth might be reshaped as a square). These maps make it easy to compare between units of very different geographical sizes, while retaining the overall rough structure of the map. This algorithm was released freely online as part of the development of the L2 free data visualisation platform.
Exploitation Route Our findings in the area of social media transport predictions could be taken forward by those who are interested in studying how other types of social data might be used to predict transport (for example, in a current piece of work funded by another grant, we are looking at whether data from open street map can forecast traffic jams). This is a very lively area of academic research. Our work on visualisations could be used by those developing broader visualisation platforms, or indeed those wishing to display data to policymakers.
Sectors Communities and Social Services/Policy,Digital/Communication/Information Technologies (including Software),Transport
URL http://smartcities.oii.ox.ac.uk/
 
Description Our work on open and social data for policy making has led to a wider partnership with Oxford City and County Councils whereby we are exploring with them different ways they can make use of data to improve policy making. For example, we are currently working with them right now on traffic data they are collecting through a partnership with a major mobility company. Our aim is to continue working collaboratively and hence build insights which can be directly used for policy making.
First Year Of Impact 2017
Sector Communities and Social Services/Policy
 
Description Google
Amount £42,892 (GBP)
Organisation Google 
Sector Private
Country United States
Start 06/2017 
End 05/2018
 
Description LRF Programme
Amount £195,176 (GBP)
Organisation Lloyd's Register Foundation 
Sector Charity/Non Profit
Country United Kingdom
Start 01/2017 
End 12/2019
 
Title Tile Map Generation Algorithm 
Description This software is an algorithm for the generation of "tile maps", which are maps where geographical objects (e.g. countries, city states, counties) are reshaped to have the same size, even though their rough position in space is preserved. This is a useful technique for displaying geographical data in a way that does not minimise the impact of small countries / states / regions etc. 
Type Of Technology New/Improved Technique/Technology 
Year Produced 2017 
Impact The paper referencing this algorithm has been cited a number of times already, and the tile maps themselves have been integrated in the L2 software display package. 
URL http://onlinelibrary.wiley.com/doi/10.1111/cgf.13200/abstract
 
Title Vizsla 
Description Vizsla is a simple JavaScript API for Vega-Lite. It improves the visualisation capabilities of this language and makes it simple to create advanced, web friendly interactive graphics. 
Type Of Technology Software 
Year Produced 2018 
Open Source License? Yes  
Impact The library is featured in Observable's visualization collection (https://observablehq.com/collection/@observablehq/visualization) and the Vega group are interested in making Vizsla the recommended API for JavaScript. 
URL https://observablehq.com/@gjmcn/vizsla-and-vega-lite