Indicative Data: Extracting 3D Models of Cities from Unavailability and Degradation of Global Navigation Satellite Systems (GNSS)
Lead Research Organisation:
University College London
Department Name: Centre for Advanced Spatial Analysis
Abstract
3-Dimensional (3D) models of cities are beneficial or even essential for many applications, including urban planning/development, energy demand/consumption modelling, emergency evacuation and responses, lighting simulation, cadastre and land use modelling, flight simulation, positioning and navigation (particularly for autonomous cars in urban canyons and disabled users requiring accessibility), and Building Information Modelling (BIM). Despite the importance of the 3D models, they are not available or being updated frequently for many areas/cities. This can be due to the process of generating and updating (by current technologies such as LiDAR (Light Detection and Ranging)) being computationally and financially expensive, time-consuming, and requiring frequent updates due to the dynamic nature of cities. This fellowship will propose and implement a crowdsourcing-based approach to create accurate 3D models from the free to use and globally available data of Global Navigation Satellite Systems (GNSS). The effects of urban features, such as buildings and trees, on GNSS signals, i.e. signal blockage and obstruction, and attenuation, will help to recognise the shape, size, and materials of urban features, through the application of statistical, machine learning (ML) and artificial intelligence (AI) techniques. The use of freely accessible raw GNSS data, which can be accessed on any current Android device, will enable the production of up to date 3D models at no or low cost, of particular value in developing regions where these models are not currently available.
GNSS is the most widely used positioning technique because of free-to-use, privacy-preserving, and globally available signals. However, GNSS signals can be blocked, reflected and/or attenuated by objects, e.g. trees, buildings, walls and windows. While blockage, attenuation and reflection of GNSS signals are common in urban canyons and indoors, making the positioning unreliable, inaccurate or impossible, the affected received signals can act as an indicator of the structure of the surrounding environments. This means, for example, if the signals are blocked or attenuated, then the size and shape of the obstacles or the type of media/material the signals have gone through or been reflected by can be understood. This needs the precise locations of satellites, and the receiver, and also predicted signal strength level at each location and time. The crowdsource-based framework, i.e. a mobile app for data capture and a web mapping application for upload of GNSS raw data, will allow the project to have well-distributed data both in space and time. This will ultimately lead to higher quality (more spatially and temporally accurate, complete, precise) 3D models. However due to the complexity of data, as neither the receiving mobile devices nor the broadcasting satellites are fixed, some novel data mining techniques, based on already existing statistical, ML, and AI techniques, need to be developed during this fellowship. They will handle the high volume, the velocity of change, and the complexity of the spatio-temporal GNSS raw data with high levels of veracity. The spatio-temporal patterns will be used for creating and updating the 3D models of cities at a high level of detail (LoDs), i.e. approximating the façade and the building materials, e.g. windows, from which the signals are reflected or have gone through. The 3D models will feed into 3D-mapping aided GNSS positioning (and integrated with other signals e.g. WiFi) which can ultimately provide more continuous and accurate GNSS positioning in urban canyons and indoors.
This fellowship will provide a novel perspective which perceives lack and degradation of data as an "indicative" source of data, which can be re-applied by other disciplines. The success of this fellowship will help me to establish myself as an internationally recognised leader in the area of spatial data science.
GNSS is the most widely used positioning technique because of free-to-use, privacy-preserving, and globally available signals. However, GNSS signals can be blocked, reflected and/or attenuated by objects, e.g. trees, buildings, walls and windows. While blockage, attenuation and reflection of GNSS signals are common in urban canyons and indoors, making the positioning unreliable, inaccurate or impossible, the affected received signals can act as an indicator of the structure of the surrounding environments. This means, for example, if the signals are blocked or attenuated, then the size and shape of the obstacles or the type of media/material the signals have gone through or been reflected by can be understood. This needs the precise locations of satellites, and the receiver, and also predicted signal strength level at each location and time. The crowdsource-based framework, i.e. a mobile app for data capture and a web mapping application for upload of GNSS raw data, will allow the project to have well-distributed data both in space and time. This will ultimately lead to higher quality (more spatially and temporally accurate, complete, precise) 3D models. However due to the complexity of data, as neither the receiving mobile devices nor the broadcasting satellites are fixed, some novel data mining techniques, based on already existing statistical, ML, and AI techniques, need to be developed during this fellowship. They will handle the high volume, the velocity of change, and the complexity of the spatio-temporal GNSS raw data with high levels of veracity. The spatio-temporal patterns will be used for creating and updating the 3D models of cities at a high level of detail (LoDs), i.e. approximating the façade and the building materials, e.g. windows, from which the signals are reflected or have gone through. The 3D models will feed into 3D-mapping aided GNSS positioning (and integrated with other signals e.g. WiFi) which can ultimately provide more continuous and accurate GNSS positioning in urban canyons and indoors.
This fellowship will provide a novel perspective which perceives lack and degradation of data as an "indicative" source of data, which can be re-applied by other disciplines. The success of this fellowship will help me to establish myself as an internationally recognised leader in the area of spatial data science.
Planned Impact
The success of this project will enable several disciplines and sectors to create, update, and (re-)use 3D models of the cities only requiring GNSS raw data, which is globally available and free-to-use. Such disciplines and sectors may include:
Energy: Currently there are several projects which run a 3D building stock model to estimate and predict energy consumption. Such 3D models are mostly based on expensive LiDAR and Ordnance Survey Mastermap. This project will provide up-to-date 3D models that are free-to-create and easy to update.
Positioning and Navigation: The primary application of GNSS is for positioning and navigation. While GNSS is the most-widely used a positioning technology with an estimated market of at £2.3B in the UK, its functionality is mainly limited to outdoors. While 3D-mapping aided GNSS positioning provides very impressive results, its accuracy and usability depend on the availability and the quality of the 3D models. This project will develop a recursive platform where GNSS RSS patterns will feed into 3D modelling and 3D models will feed into "shadow-matching" positioning services. It is also envisaged that positioning service can be integrated with 3D modelling and provides with a Simultaneous Localisation and Mapping (SLAM) service, based on GNSS raw data.
The indoor positioning and navigation applications currently are based on other positioning technologies such as Wireless Local Area Network (WLAN), Radio Frequency IDentification (RFID), Cameras, Bluetooth Low Energy (BLE), etc., none can yet provide a globally available, privacy preserving service at minimum cost to users associated with infrastructure development/installation and maintenance, or in another word a "GNSS-like" service. The success of the project will provide a seamless (indoor/outdoor) positioning and navigation service, resulting in UK industry being at the forefront of the indoor positioning market.
The seamless positioning service the project can provide, will be beneficial not only for navigation services, i.e. the biggest revenue generator of Location Based Services (LBS) industry, but also for many other services including healthcare services worldwide will benefit from a continuous, accurate positioning service improving the ability to respond quickly to emergencies, saving or improving the quality of many lives. The US statistics state 10,000 lives per year could be saved if accurate (50m horizontally) indoor location was attached to 67% of emergency calls.
Policy makers: The outcomes of this research can be utilised by policy makers (via UCL Public Policy team, Uber, OS), such as urban planners and disaster managers. E.g. with a facility to rapidly and freely generate/update the 3D model of the cities they can have a better understanding of the current situation (e.g. disaster) to manage it better. This also promotes UK involvement with European GNSS, Galileo, or mandating revisions to positional requirements for emergency calls, security and tracking related applications, in order to improve services (e.g. E112).
Public: The crowdsourcing is at the heart of the project (promoted by public engagement events and school visits). The public will contribute data and will see the 3D model of their surrounding. individuals will benefit from accessing more accurate positioning/LBS services, e.g. navigation, inside buildings.
Research: The step-changing view of this project, i.e. considering lack of data (e.g. blockage of signals) as indicative data itself (e.g. size and shape of the blocking building) can be applied by many other disciplines, such as statistics and data science. This is particularly important in the era of big data, where data might not be captured for the specific use of the application and so the level of availability and uncertainty could vary. Also, the disciplines including data science, positioning and navigation, transportation, energy, and citizen science can extend their research.
Energy: Currently there are several projects which run a 3D building stock model to estimate and predict energy consumption. Such 3D models are mostly based on expensive LiDAR and Ordnance Survey Mastermap. This project will provide up-to-date 3D models that are free-to-create and easy to update.
Positioning and Navigation: The primary application of GNSS is for positioning and navigation. While GNSS is the most-widely used a positioning technology with an estimated market of at £2.3B in the UK, its functionality is mainly limited to outdoors. While 3D-mapping aided GNSS positioning provides very impressive results, its accuracy and usability depend on the availability and the quality of the 3D models. This project will develop a recursive platform where GNSS RSS patterns will feed into 3D modelling and 3D models will feed into "shadow-matching" positioning services. It is also envisaged that positioning service can be integrated with 3D modelling and provides with a Simultaneous Localisation and Mapping (SLAM) service, based on GNSS raw data.
The indoor positioning and navigation applications currently are based on other positioning technologies such as Wireless Local Area Network (WLAN), Radio Frequency IDentification (RFID), Cameras, Bluetooth Low Energy (BLE), etc., none can yet provide a globally available, privacy preserving service at minimum cost to users associated with infrastructure development/installation and maintenance, or in another word a "GNSS-like" service. The success of the project will provide a seamless (indoor/outdoor) positioning and navigation service, resulting in UK industry being at the forefront of the indoor positioning market.
The seamless positioning service the project can provide, will be beneficial not only for navigation services, i.e. the biggest revenue generator of Location Based Services (LBS) industry, but also for many other services including healthcare services worldwide will benefit from a continuous, accurate positioning service improving the ability to respond quickly to emergencies, saving or improving the quality of many lives. The US statistics state 10,000 lives per year could be saved if accurate (50m horizontally) indoor location was attached to 67% of emergency calls.
Policy makers: The outcomes of this research can be utilised by policy makers (via UCL Public Policy team, Uber, OS), such as urban planners and disaster managers. E.g. with a facility to rapidly and freely generate/update the 3D model of the cities they can have a better understanding of the current situation (e.g. disaster) to manage it better. This also promotes UK involvement with European GNSS, Galileo, or mandating revisions to positional requirements for emergency calls, security and tracking related applications, in order to improve services (e.g. E112).
Public: The crowdsourcing is at the heart of the project (promoted by public engagement events and school visits). The public will contribute data and will see the 3D model of their surrounding. individuals will benefit from accessing more accurate positioning/LBS services, e.g. navigation, inside buildings.
Research: The step-changing view of this project, i.e. considering lack of data (e.g. blockage of signals) as indicative data itself (e.g. size and shape of the blocking building) can be applied by many other disciplines, such as statistics and data science. This is particularly important in the era of big data, where data might not be captured for the specific use of the application and so the level of availability and uncertainty could vary. Also, the disciplines including data science, positioning and navigation, transportation, energy, and citizen science can extend their research.
Organisations
- University College London (Lead Research Organisation)
- Alan Turing Institute (Collaboration)
- National University of Ireland, Maynooth (Collaboration)
- Ordnance Survey (Collaboration)
- Leibniz Institute of Ecological Urban and Regional Development (Collaboration)
- Ordnance Survey (Project Partner)
- The Alan Turing Institute (Project Partner)
- BIM Academy (Enterprises) Ltd (Project Partner)
- Leibniz Association (Project Partner)
- University of Nottingham (Project Partner)
- University of Glasgow (Fellow)
Publications
Shin H
(2022)
The Impact of Built Environment on Bike Commuting: Utilising Strava Bike Data and Geographically Weighted Models
in AGILE: GIScience Series
Lines T
(2021)
3D map creation using crowdsourced GNSS data
in Computers, Environment and Urban Systems
Yan J
(2020)
3-D Passive-Vision-Aided Pedestrian Dead Reckoning for Indoor Positioning
in IEEE Transactions on Instrumentation and Measurement
Basiri A
(2019)
Crowdsourced geospatial data quality: challenges and future directions
in International Journal of Geographical Information Science
Basiri A
(2020)
Navigating Through Pandemic: The Use of Positioning Technologies
in Journal of Navigation
Basiri A
(2021)
Predictably unpredictable
in Journal of Navigation
Basiri A
(2021)
Inclusivity and diversity of navigation services
in Journal of Navigation
Basiri A
(2021)
How Fast Can Our Horses Go? Measuring the Quality of Positioning Technologies
in Journal of Navigation
Shubina V
(2021)
Effectiveness modelling of digital contact-tracing solutions for tackling the COVID-19 pandemic
in Journal of Navigation
Basiri A
(2021)
'A novel model blah blah blah'
in Journal of Navigation
Mitra R
(2023)
Learning from data with structured missingness
in Nature Machine Intelligence
Basiri A
(2022)
Missing data as data.
in Patterns (New York, N.Y.)
Lines T
(2020)
Signal Attenuation Modelling in WLAN Positioning
Description | We have already proven that the estimation of height and some types of materials of the buildings from the blockage, reflection, and attenuation of GPS (and in general Global Navigation Satellite Systems (GNSS)) data. We found we can extract some patterns using statistical, machine learning (ML), techniques applied to semi- crowdsourced GNSS raw data, contributed by the volunteers. This will ultimately provide us with a ubiquitous and free of charge 3D models creation/update. |
Exploitation Route | two main groups of (a) the research communities that benefit from or study the 3D modelling service, and (b) the research communities that can reapply the novel data science methodologies and/or the mindset of "indicative data science", i.e. extracting data/information from lack/degradation of data. The first groups of disciplines and research communities include energy demand/consumption modelling, smart cities, cadastre and land use modelling, (3D-map aided GNSS) positioning and navigation, intelligent mobility and transport, and Building Information Modelling (BIM), lighting simulation. They can (i) apply the results to achieve better service they already provide using 3D maps that are generated/purchased by conventional techniques using publically available and free to use 3D modelling service. They can also (ii) further this research by studying the quality and fitness-for-purpose of the created 3D maps, the requirements of the input data, i.e. minimum number of the receivers (spatial density), temporal intervals of making observations at each location, geometry of the receivers for different Level of Details (LoDs) and different applications and scenarios of use. Also some services that use 2D maps currently, e.g. navigation and routeing, can study the impacts of using 3D maps of the quality of service. The second group of disciplines and research communities include data science, Machine Learning (ML), Artificial Intelligence (AI), Volunteered Geographic Information (VGI) and citizen science, social media and big data analytics. They can modify and/or extend on the to-be-developed novel techniques customised for other indicative datasets. This will be made possible through open and re-producible programming platforms and release of libraries and APIs. Many disciplines that are using indicative data such as social media and VGI that can have some levels of bias, availability, and uncertainty, can extend and be reapplied the ML, AI and big data analytics for such datasets. Lack of high quality (spatially, temporally and thematically accuracy, consistent, and complete) data can result in low-quality output if conventional techniques are applied. The extension of this spatio-temporal patterns and mining techniques, which are novel statistical, ML, and AI techniques addressing the GIGO and embedded biases in the input data can be studied by many other disciplines. |
Sectors | Aerospace Defence and Marine Construction Digital/Communication/Information Technologies (including Software) Electronics Energy Environment Culture Heritage Museums and Collections Transport |
URL | https://www.indicativedatascience.com/ |
Description | The project developed in a globally available, free-to-use, 3D modelling service based on crowdsourced data, which will require no additional infrastructure. This will provide a particular impact in developing nations where 3D modelling has previously been prohibitively expensive. Also as a crowdsourced-based project, the engagement of volunteers will be a core element of the design. The public involvement with GNSS raw data will increase the general understanding of GNSS, and 3D mapping. Societal Impacts: A lar part of the team is working on citizen science and public participatory aspects of the project to ensure the crowdsourcing app (and in general any crowdsourcing project) is truly available to the crowd and not technically only. This will help the issue of the biased/long-tail population of the "crowd" in new forms of data. Working with human factor team, Alan Turing's Public policy programme and Uber will help more user-friendly web and mobile app interface design, and better public engagement plans. Several online training courses, project webpage updates, and news and media releases are already released. This include a new special interest group looking specifically into the issues and challenges of not having ALL in AI: https://www.turing.ac.uk/research/interest-groups/facilitating-responsible-participation-data-science These will be used as vehicles to train the public volunteers. I, as a registered STEM ambassador, will organise three workshops and school visits with children in two schools, and will use my experience in outreach activities, including my work as an active member of Women In Science and Engineering, and EU Role Model in Science to organise interactive events. These will expose students and the public to engineering and explain the benefits of participation and of the outputs of the research to inspire the next generation of STEM students and to learn more about the crowdsourcing, 3D maps, and GNSS. Press releases and a project website will be used to disseminate research outputs and engage with public and potential additional collaborators. In addition, the output will improve several public services including transport, healthcare, emergency and security services, which require 3D models of cities and/or accurate positioning and navigation in dense urban canyons. This is particularly important from a societal perspective, not only industrial, as such public services greatly improve the quality of lives of citizens. For example, healthcare services worldwide will benefit from a continuous, accurate 3D-map aided GNSS positioning service. This will improve the ability to respond quickly to emergencies, saving or improving the quality of many lives. US Federal Communication Commission statistics state 10,000 lives per year could be saved if accurate (50m horizontally) indoor location was attached to 67% of emergency calls. The outcomes of this research can be utilised by policy makers, for example, by mandating revisions to positional requirements for emergency calls, security and tracking related applications, in order to improve emergency services (e.g. E112). A report on the achievable quality of service will be published on the website and disseminated through media press releases informing the public and policymakers of such assessment. This report will influence policymakers and government agencies, local government, the Department of Transport, and so shape future policy decisions. This will be promoted by and made available to the government Satellite Applications Catapults, Future Cities Catapults to maximise the impacts. One event will engage policymakers (government agencies, local government, Department of Transport, Transport for London, and industrial partners and their contacts Academic Impacts: The crowdsourced, open, and free 3D modelling service will benefit several disciplines and research projects that use 3D maps/models for their research. They include energy demand/consumption modelling, smart cities, cadastre and land use modelling, (3D-map aided GNSS) positioning and navigation, intelligent mobility and transport, and Building Information Modelling (BIM), lighting simulation. They can (i) apply the results to achieve better service they already provide using 3D maps that are generated/purchased by conventional techniques using publicly available and free to use 3D modelling service. They can also (ii) further this research by studying the quality and fitness-for-purpose of the created 3D maps, the requirements of the input data, i.e. minimum number of the receivers (spatial density), temporal intervals of making observations at each location, geometry of the receivers for different Level of Details, applications, and scenarios. Also, some services that are currently using 2D maps, e.g. navigation, can benefit from 3D maps of the quality of service. (iii) In addition to direct and immediate academic impacts, several research communities can re-apply the novel data science methodologies and/or the mindset of "indicative data science", i.e. extracting data/information from lack/degradation of data. In the era of big data, social media and crowdsourcing, when data might not be captured for the specific use of the application and so data may have some levels of bias, availability, and uncertainty, the perspective of indicative data, can be re-applied, customised, and extend by many disciplines. As a RAEng EngineeringX I am actively involved with the public on the environmental and legal aspects of storage of Digital Data beyond we need. Digital Data, unlike most phenomena, do not have a finite lifetime, and technically speaking can be stored forever. Such digital 'immortality' introduces legal, environmental, security, and safety challenges. For example, data protection regulations, including GDPR, may not protect our personal data after our physical death. While our digital data can outlive ourselves, the tech companies get to decide how to deal with our data and accounts. Digital inheritance is not recognised in many countries, despite the rise of cryptocurrencies. Also, data storage and maintenance are energy-demanding, and not setting an end-of-lifetime for digital data can have environmental impacts. I discuss with the public, tech companies, and policymakers about the issues associated with not having an end-of-lifetime for digital data set at the beginning, including the environmental and legal issues. I engage with public and policymakers to raise awareness about the need for digital inheritance legislations. I communicate the environmental aspects and the carbon footprint of our digital lifestyle with the public and "digital is not always greener". I work with tech companies and developers to implement a transparent data retention plan and to include the digital death statements in the consent forms from for the users when sign-up. Economic Impacts: The free and globally available services, that only require GNSS raw data, to provide 3D city models, and also 3D-map aided GNSS positioning can open up and/or enhance several commercialised applications and services. GNSS and LBS markets are a major growth area, but mainly limited to outdoor applications. There is additional growth potential associated with the availability of accurate indoor positioning systems, with UK markets estimated at £2.3B. The project success will result in UK industry being at the forefront of the indoor positioning and navigation market. Two events will focus on the design of the pilot scenarios/applications with collaborators from OS, BIM Academy, Uber Satellite Application Catapult, and Future Cities Catapults. I will work closely with contract and business team at ATI and Glasgow to develop links with the public and businesses and to commercialise outputs where appropriate. This will include working with them to put in place Confidentiality Agreements and advice on IP issues and potential patent applications. Environmental Impacts: The free and ubiquitous 3D models will enable energy consumption modelling to estimate and model users more accurately. They currently run 3D building stock models, mostly based on expensive LiDAR and OS Mastermap. This project will provide up-todate 3D models that are free-to-create and easy to update. In addition, more continuous 3Dmap aided positioning, with the help of Uber will help traffic and transport management, resulting in less congested areas, due to more accurate navigation services, and so can eventually reduce the emission of carbon dioxide and improve local air quality and health outcomes. |
First Year Of Impact | 2020 |
Sector | Construction,Digital/Communication/Information Technologies (including Software),Education,Environment,Transport |
Impact Types | Cultural Societal Economic Policy & public services |
Description | public policy debate, and media engagement |
Geographic Reach | National |
Policy Influence Type | Participation in a guidance/advisory committee |
Impact | I actively attended public policy debate through active panel discussions (e.g. Benchmark action: https://vimeo.com/376826396?ref=tw-share) and media engagement, such as the interviews with TechWorld: https://www.techworld.com/data/mapping-location-data-ethics-of-where-we-are-where-we-could-be-3779814/) on location privacy and role of EU's GDPR which should be adopted after Brexit. |
URL | https://www.techworld.com/data/mapping-location-data-ethics-of-where-we-are-where-we-could-be-377981... |
Title | GnssMapper (3D Mapping using GNSS Signals) |
Description | GnssMapper provides tools for generating 3D maps by using Global Navigation Satellite System (GNSS) data. It is the results of UKRI Future Leaders Fellowship research project at the University of Glasgow, which investigates methods for using crowdsourced GNSS data for mapping. This provides a novel approach to generate 3D maps for free using Global Navigation Satellite Systems (GNSS) signals, which are globally available. The Results of the proposed technique implemented by GnssMapper demonstrate the ability to create 3D maps using automatically processed GNSS data. The results show that the third dimension, i.e. height of the buildings, can be estimated with below 5m accuracy, which is comparable to the accuracy of the national mapping agency of Great Britain, Ordnance Survey GB. 3D maps are increasingly useful for many applications such as drone navigation, emergency services, and urban planning. However, creating 3D maps and keeping them up-to-date using existing technologies, such as laser scanners, is expensive. This software implements a novel approach to generate 2.5D (otherwise known as 3D level-of-detail (LOD) 1) maps for free using Global Navigation Satellite Systems (GNSS) signals, which are globally available and are blocked only by obstacles between the satellites and the receivers. This enables us to find the patterns of GNSS signal availability and create 3D maps. It applies algorithms to GNSS signal strength patterns based on a boot-strapped technique that iteratively trains the signal classifiers while generating the map. Results of the proposed technique demonstrate the ability to create 3D maps using automatically processed GNSS data. The results show that the third dimension, i.e. height of the buildings, can be estimated with below 5 metre accuracy, which is the benchmark recommended by the CityGML standard. GnssMapper is written in Python and built upon GeoPandas objects. It provides the following capabilities: -read 'raw' GNSS data from Google's gnsslogger app, available for Android phones -process data into a set of observations -estimate building heights based on the observations -simulate observations for algorithm testing -It does not include any functionality for processing GNSS data in order to estimate position, and assumes position data is available from the log file, or calculated elsewhere. GnssMapper depends on the GeoPandas package and its underlying dependencies, including PyGeos. You may view the source code at https://github.com/Indicative-Data-Science/gnssmapper You may read the details of the research on 3D mapping using GNSS here https://www.sciencedirect.com/science/article/pii/S0198971521000788 GnssMapper has been used by three MSc dissertations in Imperial College London, University of Glasgow, University College London, and currently is used with several PhD students in different countries including China, UK, US and South Africa. |
Type Of Material | Improvements to research infrastructure |
Year Produced | 2021 |
Provided To Others? | Yes |
Impact | GnssMapper provides tools for generating 3D maps by using Global Navigation Satellite System (GNSS) data. It is the results of UKRI Future Leaders Fellowship research project at the University of Glasgow, which investigates methods for using crowdsourced GNSS data for mapping. This provides a novel approach to generate 3D maps for free using Global Navigation Satellite Systems (GNSS) signals, which are globally available. The Results of the proposed technique implemented by GnssMapper demonstrate the ability to create 3D maps using automatically processed GNSS data. The results show that the third dimension, i.e. height of the buildings, can be estimated with below 5m accuracy, which is comparable to the accuracy of the national mapping agency of Great Britain, Ordnance Survey GB. 3D maps are increasingly useful for many applications such as drone navigation, emergency services, and urban planning. However, creating 3D maps and keeping them up-to-date using existing technologies, such as laser scanners, is expensive. This software implements a novel approach to generate 2.5D (otherwise known as 3D level-of-detail (LOD) 1) maps for free using Global Navigation Satellite Systems (GNSS) signals, which are globally available and are blocked only by obstacles between the satellites and the receivers. This enables us to find the patterns of GNSS signal availability and create 3D maps. It applies algorithms to GNSS signal strength patterns based on a boot-strapped technique that iteratively trains the signal classifiers while generating the map. Results of the proposed technique demonstrate the ability to create 3D maps using automatically processed GNSS data. The results show that the third dimension, i.e. height of the buildings, can be estimated with below 5 metre accuracy, which is the benchmark recommended by the CityGML standard. GnssMapper is written in Python and built upon GeoPandas objects. It provides the following capabilities: -read 'raw' GNSS data from Google's gnsslogger app, available for Android phones -process data into a set of observations -estimate building heights based on the observations -simulate observations for algorithm testing -It does not include any functionality for processing GNSS data in order to estimate position, and assumes position data is available from the log file, or calculated elsewhere. GnssMapper depends on the GeoPandas package and its underlying dependencies, including PyGeos. You may view the source code at https://github.com/Indicative-Data-Science/gnssmapper You may read the details of the research on 3D mapping using GNSS here https://www.sciencedirect.com/science/article/pii/S0198971521000788 GnssMapper has been used by three MSc dissertations in Imperial College London, University of Glasgow, University College London, and currently is used with several PhD students in different countries including China, UK, US and South Africa. |
URL | https://github.com/Indicative-Data-Science/gnssmapper |
Title | Missing/Biased Data Science |
Description | Unrepresentative data leads to flawed conclusions; a problem when decisions in industry and government are increasingly data-driven. "Big" data are likely unrepresentative but new methods are needed to understand them. Understanding (a) what is the equivalent sample size of self-reporting/crowdsourced data (such as social media account) which might be biased in a randomised survey? (b) how spatial microsimulation may be affected by survey representativeness? and (c) if anonymous behaviour data be tested for representativeness using "mass imputation", through combine datasets when items can't be directly linked. |
Type Of Material | Improvements to research infrastructure |
Year Produced | 2020 |
Provided To Others? | No |
Impact | New forms of data, in the era of big data, are increasingly available while their representativeness and completeness of data are under question. This will help to address the big data paradox in that sense (quality vs. qantity) in many areas of research. |
Title | Building height algorithm |
Description | 3-Dimensional (3D) models of cities are beneficial or even essential for many applications, including urban planning/development, energy demand/consumption modelling, emergency evacuation and responses, lighting simulation, cadastre and land use modelling, flight simulation, positioning and navigation (particularly for autonomous cars in urban canyons and disabled users requiring accessibility), and Building Information Modelling (BIM). Despite the importance of the 3D models, they are not available or being updated frequently for many areas/cities. This can be due to the process of generating and updating (by current technologies such as LiDAR (Light Detection and Ranging)) being computationally and financially expensive, time-consuming, and requiring frequent updates due to the dynamic nature of cities. This algorithm estimates the height of a building given a 2D map of the building outline and a a dataset of GNSS readings, by applying statistical, machine learning (ML) and artificial intelligence (AI) techniques to GNSS signal characteristics such as time of flight and signal strength to identify signal blockage and obstruction, and attenuation. The algorithm is currently under development, to be published later in 2020. |
Type Of Material | Computer model/algorithm |
Year Produced | 2020 |
Provided To Others? | No |
Impact | Used by our research group to show the potential use of GNSS signals to create 3D models of cities. |
Title | Database of GNSS observations |
Description | Raw GNSS data, which can be accessed on any current Android device, is a novel form of geographic data. It has high potential for large scale use in human geographic study as it is widely accessible and based on an underlying technology which is free-to-use, privacy-preserving, and globally available. The signal characteristics and accuracy of the automatically estimated location vary due to the surrounding environment which may cause blockage, attenuation and reflection of GNSS signals. Therefore study of such datasets is complicated due to the need to infer jointly on position and on the signal characteristics of interest. A database of raw GNSS readings has been collected by the research team within a small area (0.25 sq km) manually labelled with geolocations to a high degree of accuracy. The database is currently partially collected, with the intent to make the database accessible to other researchers once completed. |
Type Of Material | Database/Collection of data |
Year Produced | 2021 |
Provided To Others? | Yes |
Impact | Used by the research team to evaluate the feasibility of inferring the surrounding environment from signal characteristics, in work to be published later in 2020 and 2021. |
URL | https://github.com/Indicative-Data-Science/gnssmapper/tree/master/examplefiles |
Title | GNSS post processing in R |
Description | Raw GNSS data, which can be accessed on any current Android device, is a novel form of geographic data. Tools exist for processing GNSS data to estimate the receiver position but the majority are incompatible with the data format used by Android. Furthermore new applications for the data require the ability to access processed individual signals as opposed to the overall receiver position estimate. A set of tools have been written in R which perform standard GNSS post-processing techniques to the Android produced data. This replicates the functionality of similar tools written in other programming languages, but is compatible with the Android format, and allows the interim output to be available for further data analysis in R. |
Type Of Material | Data analysis technique |
Year Produced | 2020 |
Provided To Others? | No |
Impact | Used by our research group to evaluate the feasibility of inferring about the surrounding environment from GNSS signal characteristics. |
Description | Alan Turing Institute |
Organisation | Alan Turing Institute |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Prof Ana Basiri serves the Programmes at the Alan Turing Institute in several ways including leadership (TPS Program), joined projects and studentships (with Health, DnS, ASG), presenting ATI in different event and network, and facilitating collaboration between the Turing and Glasgow University as the Turing Network Development Awardee. |
Collaborator Contribution | The Alan Turing Institute supports Professor Ana Basiri's fellowship project and closer collaboration with the University of Glasgow which closely aligns with our Data Centric Engineering (DCE) programme, and Defence and Security Programme and will complement projects within the ATI's portfolio aligned to our strategic challenge to deliver resilient and robust infrastructure. ATI commit the following resources towards this project (estimated £20k, contribution in kind): • Senior management input to steer the research (c. 5 days / year, subject to availability) • Hosting project meetings and workshops at the Alan Turing Institute (subject to availability) • Access to the DCE community of experts to look for collaborative opportunities to advance the research |
Impact | Networking, Output (papers, data, software) meetings and presentations software packages |
Start Year | 2019 |
Description | Leibniz Institute of Ecological Urban and Regional Development |
Organisation | Leibniz Institute of Ecological Urban and Regional Development |
Country | Germany |
Sector | Private |
PI Contribution | Using the 3D mapping service we are testing the automatic building type recognition based on the 3D model, testing the automatic detection of changes in the building stock over time (e.g. demolition, front/back extensions, additional storeys, balcony extensions etc.) based on multi-temporal 3D models. |
Collaborator Contribution | IOER contributes to this project through (a) advisory activities and participation in workshops related to WP1 to WP4 to define test scenarios for the implementation of the proposed techniques, evaluation of the results and commenting on the methodology. They give advice from the perspective of urban/regional planning and sustainable development (applications, use case scenarios, etc.) but also from the methodological point of view (design of the crowdsourcing app, data quality measures). (b) they support the project with the provision of data from the IOER monitor (http://www.ioer-monitor.de/en/) and advise you in acquiring other reference data on buildings that is free of charge (Open data). We have altready (c) initiated collaborative research activities, e.g. 1-2 co-supervised master / doctoral thesis, which might lead to further collaborative research proposals. The IOER is also willing to provide research infrastructure for 1-2 CASA students / doctoral candidates visiting the IOER with access to all resources (especially data for validation) and a supervision in an interdisciplinary environment. Also, (d) testing and evaluating the quality, and the fitness-for purpose of the 3D model for different scenarios in Germany. This may include assisting the engagement of volunteers, reference data collection, and data quality analysis by comparing with official data sets/internal reference data, etc. (e) Testing our approaches on automatic building type recognition based on the 3D model and GNSS RSS degradation data and (f) Testing approaches for the automatic detection of changes in the building stock over time (e.g. demolition, front/back extensions, additional storeys, balcony extensions etc.) based on multi-temporal 3D models. |
Impact | This is a very multidisciplinary collaboration on several areas: IOER contributes to this project through (a) advisory activities and participation in workshops related to WP1 to WP4 to define test scenarios for the implementation of the proposed techniques, evaluation of the results and commenting on the methodology. They give advice from the perspective of urban/regional planning and sustainable development (applications, use case scenarios, etc.) but also from the methodological point of view (design of the crowdsourcing app, data quality measures). (b) they support the project with the provision of data from the IOER monitor (http://www.ioer-monitor.de/en/) and advise you in acquiring other reference data on buildings that is free of charge (Open data). We have altready (c) initiated collaborative research activities, e.g. 1-2 co-supervised master / doctoral thesis, which might lead to further collaborative research proposals. The IOER is also willing to provide research infrastructure for 1-2 CASA students / doctoral candidates visiting the IOER with access to all resources (especially data for validation) and a supervision in an interdisciplinary environment. Also, (d) testing and evaluating the quality, and the fitness-for purpose of the 3D model for different scenarios in Germany. This may include assisting the engagement of volunteers, reference data collection, and data quality analysis by comparing with official data sets/internal reference data, etc. (e) Testing our approaches on automatic building type recognition based on the 3D model and GNSS RSS degradation data and (f) Testing approaches for the automatic detection of changes in the building stock over time (e.g. demolition, front/back extensions, additional storeys, balcony extensions etc.) based on multi-temporal 3D models. |
Start Year | 2019 |
Description | Maynooth University |
Organisation | Maynooth University |
Country | Ireland |
Sector | Academic/University |
PI Contribution | Joint studentships with NCG, organising events, exchange of staff from the team, and join courses. |
Collaborator Contribution | the National Centre for Geocopmutation supports Prof Ana Basiri's research fellowship. The National Centre for Geocomputation (NCG) at Maynooth University en- joys a worldwide reputation in research in spatial data analytics, geographical data analysis and management, and the use of airborne sensor technology, applied in a number of areas including health, climate change, demographics and agricultural applications. The NCG offers support in a number of ways; Firstly this can involve members of the NCG attending meetings, advisory contribution to research projects, input to the supervision of PhD students and post-doctoral researchers, hosting and exchanging staff between the NCG and the School of Geographical & Earth Sciences at the University of Glasgow, collaboration on publications, and the possible hosting of joint research themed workshops or similar events. The in-kind resource value quoted below is indicative and our offer covers whatever resources are eventually required in the research, for the duration of the fellowship. We have estimated in-kind costs out to 5 years as our offering is very likely to have changed within the possible 7-year span of this fellowship. |
Impact | Joint studentship application to Scottish Graduate School of Social Science (SGSSS) funded by ESRC on good data-big-data. |
Start Year | 2020 |
Description | Ordnance Survey |
Organisation | Ordnance Survey |
Country | United Kingdom |
Sector | Public |
PI Contribution | The output of the project, free 3D mapping service, will improve the quality of OS maps (and staffing for sending surveyers to update maps, only when the crowdsourcing app finds changes). |
Collaborator Contribution | Ordnance Survey has made significant in-kind contributions to the fellowship project through staff time and the supply of data, both OS products and in-house 'research data'. The in-kind data value quoted below is indicative and our offer covers whatever geographic extent of data is eventually required in the research. |
Impact | Future plans for partially funding Ph.D. studentships, joined seminars and workshop hosted/endorsed by OS. |
Start Year | 2019 |
Title | GNSSMapper |
Description | GnssMapper provides tools for generating 3D maps by using Global Navigation Satellite System (GNSS) data. |
Type Of Technology | Software |
Year Produced | 2021 |
Open Source License? | Yes |
Impact | GnssMapper provides tools for generating 3D maps by using Global Navigation Satellite System (GNSS) data. It is connected to a research project at the University of Glasgow, which investigates methods for using crowdsourced GNSS data for mapping. It is written in Python and built upon GeoPandas objects. It provides the following capabilities: read 'raw' GNSS data from Google's gnsslogger app, available for Android phones process data into a set of observations estimate building heights based on the observations simulate observations for algorithm testing It does not include any functionality for processing GNSS data in order to estimate position, and assumes position data is available from the log file, or calculated elsewhere. |
URL | https://github.com/Indicative-Data-Science/gnssmapper |
Description | Debate and podcast |
Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Policymakers/politicians |
Results and Impact | I served a panel on social distancing technologies, which was streamed on https://vimeo.com/424701542. I sat with Zoe Tabary, Property Rights Editor, Thomson Reuters Foundation, Andrew Pakes, Director of Communications & Research, Prospect, Gina Neff, Associate Professor at the Oxford Internet Institute and the Department of Sociology, University of Oxford. Author Venture Labor (2012), co-author Self-Tracking (2016), Leo Scott Smith, Founder and CEO, Tended and discuss the possibilities, risks and ethics of tracking employees' movement to reduce risk of infection from the coronavirus. What are the rights and responsibilities of employers and employees around location tracking at work? How do the technologies and devices work and how widely are they being used? What are the main benefits and risks associated with them, and how could they be improved? What can employees require, as assurance of safe conditions? And as oversight / partnership in the system? Given concerns that tracking can also be used to monitor and drive productivity, how can employers gain the trust of employees? |
Year(s) Of Engagement Activity | 2020 |
URL | https://vimeo.com/424701542 |
Description | Interview |
Form Of Engagement Activity | A magazine, newsletter or online publication |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Public/other audiences |
Results and Impact | I have contributed to: -public policy debate through active panel discussions (e.g. Benchmark action: https://vimeo.com/376826396?ref=tw-share), -media engagement, such as interview with TechWorld (https://www.techworld.com/data/mapping-location-data-ethics-of-where-we-are-where-we-could-be-3779814/), -public release of Royal Institute of Navigation ( https://twitter.com/at_RIN/status/1234779319693651968 and https://rin.org.uk/page/NavigationNews), Alan Turing Institute champagne for Women in Data Science (https://www.youtube.com/watch?v=P2N78tKwQxM&feature=youtu.be and https://twitter.com/turinginst/status/1236957027584401408) and UCL (https://www.ucl.ac.uk/bartlett/casa/news/2019/jan/personalised-solutions-better-navigation-services). |
Year(s) Of Engagement Activity | 2019,2020 |
URL | https://www.techworld.com/data/mapping-location-data-ethics-of-where-we-are-where-we-could-be-377981... |
Description | Invited talk |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Third sector organisations |
Results and Impact | Gave a talk at Alan Turing Inst's Urban Analytics program on missing data: https://www.turing.ac.uk/sites/default/files/2020-06/ana_basiri_-_presentation.pdf |
Year(s) Of Engagement Activity | 2020 |
URL | https://www.turing.ac.uk/events/urban-analytics-monthly-meet |