Posters Galore: 'BT Order Journey Data'

Mahsa Abazari Kia
University of Essex

Many companies such as BT have a significant workforce of field engineers, desk-based agents and customer support services who collect and manage large volumes of unstructured and semi-structured information.

Common to this information is that it usually comes as free-form text that is difficult to access.

The goal of the project is to unravel this unstructured information and to create business value by the automated extraction and summarization of information.

We develop novel methods for automated text analyses and summarization using modern Deep Learning techniques. We extract relevant entities and events from the text and contextualize this information into human-readable summaries.

This will enable us to exploit the information for new services of BT.

View the poster here
View more posters and videos

Posters Galore: 'Exploiting Mixed-Reality Technologies for Interacting with BIM models to Assist Onsite Activities'

Hugo Leon-Garza
University of Essex

Building Information Modelling is a cooperative process for creating and managing information on a construction project.

As a result of this process, Building Information Models (BIM models) are obtained - these models are data-rich, object-oriented, intelligent, and parametric digital representations containing all the information from all the stakeholders of a facility or building. The elements in the model have properties and relationships which can be exploited by utility companies in planning and maintenance activities.

Being capable of bringing this kind of information to onsite activities is expected to improve the quality of the field engineer’s work.

Augmented Reality (AR), is a Mixed-Reality technology that supplements the reality by superimposing digital information in the user’s view.

Utility companies, such as British Telecom (BT), can use this technology to assist their engineers with all the information related to a location. However, BIM are complex and size heavy data objects, which make it difficult for mobile and wearable hardware with limited performance capabilities, such as augmented reality glasses, to load the complete model.

In this work we show how we could process a BIM model and make use of a Mixed-Reality headset (Microsoft HoloLens), to visualize and interact with the elements. We also propose a Fuzzy Logic approach to load objects to the user’s view by batches in a way that the application performance is minimally affected

• View the poster here
• View more posters and videos

Posters Galore: Securing Software Defined Networks Through Blockchain Technology

Dr Jiejun Hu
University of Essex

The Internet of Things (IoT) is coming fast upon us. With its arrival, industry, homes, and society will potentially suffer from increased security risks through the large volume of connected IoT devices. This could overturn the huge benefits of this forthcoming cyber-physical reality.

SerIoT is a project funded by the EUs Horizon 2020 Research and Innovation programme (grant No 780139) that addresses the information security in IoT platforms and networks.

It works in a holistic, cross-layered manner i.e. through solutions in IoT platforms & devices, honeypots, software defined networking (SDN). Within SerIoT, this work proposes a blockchain-based SDN framework to provide a novel solution of the potential security issues in SDN.

Specifically, we propose a blockchain-as-a-service (BaaS) framework to support:

1. switch flow verification and insertion
2. secure communication between a SDN switch and controller
3. protection of the connection privacy among the IoT devices
4. fast deployment of blockchain for SerIoT.

Use of an external BaaS may lead to lack of knowledge of the blockchain agents who are responsible for new flow verification/validation by running a smart contract.

 

We need to design a strategy to prevent the blockchain agents from acting arbitrarily, as this would result in what is termed a Moral hazard. Thus, this work develops a novel mathematical model of the fair reward scheme through game theory. The simulation results demonstrate that the proposed algorithm balances the needs of the blockchain agents to maximise the overall social welfare, i.e. the sum of profit across all parties.

 

• View the poster here
• View more posters and videos

Posters Galore: 'A Statistical Framework for the UK’s Mobile Network'

Dave Townend
University of Essex

A key challenge in researching new wireless cellular technologies and architectures is accurate modelling of their impact on a large scale network. The following research aims to provide a statistical framework to characterise a real national mobile network such that the researcher community can develop representative models and simulations to evaluate new wireless technologies.

Current industry guidelines relevant to modelling of wireless systems and their deployments, including those defined by standards bodies such as ITU and 3GPP, are often over simplified and based on generalisation of the network environment and topology.

This research details statistical deployment characteristics of mobile base station sites and their associated geographic environment based on a highly detailed 3D topology model of large areas of the UK. High resolution LIDAR (Light Detecting and Ranging) data and cell mast properties are utilised to generate statistical descriptions of urban, suburban and rural cell site types and distributions of the building and ‘clutter’ properties that can be found within each cell type’s wireless coverage area

• View the poster here.
• View more posters and videos

Posters Galore: 'Wideband RF Over Fibre/IP'

Matthew Evans
University of Essex, Aber Electronics

The outside broadcast industry is built upon the successful transmission and reception of data modulated onto RF signals. Technological advancements have not been experienced in the same way as in other industries; solutions simply rehash old ideas. Some applications have moved towards the use of optical fibre transmission mediums for RF signals bolted onto existing installations; this project was targeted towards a new integrated adaptable solution.

The project was a Knowledge Transfer Partnership (KTP) collaboration between ABER Electronics Ltd. and the University of Essex; partially funded by Innovate UK.

The first stage of the project was to design and develop a highly versatile downconverter, integrated with an improved custom antenna design and an existing radio-over-fibre (RoF) transmitter. The RoF transmitter (and subsequently receiver) are existing products on the market so a key partnership was setup to allow for easier integration and support of the system. The second stage of the project was to develop a high bandwidth bi-directional RF over IP system using a software defined radio (SDR) module. The third, and final, stage was to redesign the existing range of power amplifiers offered by the host company; to significantly improve their operating efficiencies.
This has since morphed into a full-blown PhD research project with exciting ideas and developments being explored. Future work is to be focussed around the completion of the green amplifiers research PhD as well as developing a fully end-to-end solution for the outside broadcast industry with one eye on advancements made in the 5G research area.

View the poster here.
View more posters and videos

Thought Leadership Event: Russell Beck of ImagineThinkDo on 'The Future of Work'

Thought Leadership Event
Speaker: Russell Beck, ImagineThinkDo
Title: The Future of Work

Russell Beck, ImagineThinkDo, recently presented our latest Applied Research & Tommy Flowers Network thought leadership event.

Focusing on “The Future of Work”, Russell took us on a journey exploring global megatrends playing out in the world and what that means to people. The important question, so what? what do these trends mean to work and our future lives as employees?

Russell covered seven points:  Work Nomads; Technology; Women; Future Skills; Be the Best, Brand and Networking.

What really came across was how to build your strengths and skills; building your own personal brand and how important networking is.

Russel was one of our speakers at the conference on 'Creating Value for Economies: Collaborate - Innovate - Educate', which we had to cancel due to the COVID-19 crisis. Please have a look at his speaker profile for more information about him and his company.

Posters Galore: 'Smarter Suffolk: sensors and data for public services'

Dr Hannah Steventon and Prof Nicholas Caldwell
University of Suffolk

As climate change necessitates energy reduction and public funds are stretched, increasingly affordable Internet of Things (IoT) sensors and cutting-edge Low Power Wide Area Networks, are changing how Suffolk delivers public services. Dr Hannah Steventon and Professor Nicholas Caldwell, from University of Suffolk, describe how the University is working with Suffolk County Council (SCC) by trialling hundreds of IoT sensors and connecting them via a new Suffolk-wide LoRa network.  The aim, to improve targeting of street lighting, winter gritting, and drainage cleaning. 

 

Showcase trials at the new UoS DigiTech centre at Adastral Park are comparing sensors across price-points - from manufacturers based in East Anglia and further afield - measuring road surface temperature, ambient lighting levels, air quality and gully operation.

 

Two case studies focus on Street Lights in Suffolk and Gritting Suffolk roads. 

• Sixty-five thousand street lights in Suffolk have an energy cost of 2.6 million each year, with associated environmental impact. SCC is continuing its leading role in energy-saving adaptive street lighting with real-time dimming of lighting on arterial routes, based on traffic-counts and ambient light levels. UoS data analysis of traffic use data, from radar around Ipswich, is increasing the understanding of road usage patterns and identification of unusual road use. 
• Gritting over two thousand miles of Suffolk roads is a major operation each county-wide gritting action. Measurement of road surface temperature at closer intervals could enable targeted gritting. IoT road surface temperature sensors from three manufacturers are being benchmarked and their results compared for veracity and reliability of temperature data. Evaluating sensors and analysing data inform meaningful cost-benefit decisions and implementation plans, therefore enabling SCC to deliver a Smarter Suffolk.

Hannah would be delighted for further discussion, comments and questions at h.steventon@uos.ac.uk

 

• View the poster here
• View more posters and videos