Skip to main

Impact case study database

The impact case study database allows you to browse and search for impact case studies submitted to the REF 2021. Use the search and filters below to find the impact case studies you are looking for.

Search and filter

Filter by

  • The University of Essex
   None selected
  • 11 - Computer Science and Informatics
   None selected
   None selected
   None selected
   None selected
   None selected
   None selected
Waiting for server
Download currently selected sections for currently selected case studies (spreadsheet) (generating)
Download currently selected case study PDFs (zip) (generating)
Download tags for the currently selected case studies (spreadsheet) (generating)
Currently displaying text from case study section
Showing impact case studies 1 to 6 of 6
Submitting institution
The University of Essex
Unit of assessment
11 - Computer Science and Informatics
Summary impact type
Technological
Is this case study continued from a case study submitted in 2014?
No

1. Summary of the impact

Essex research for the rail industry resulted in the world's first 10GbE Ethernet train backbone over copper; achieving bandwidth capacity previously considered impossible. The 10GbE backbone was developed using rail-approved copper cable technology in collaboration with LPA Connection Systems, transforming the company’s fortunes and propelling the company to become the market leader [text removed for publication]. Since the technology’s launch LPA’s turnover increased by 60% with 10GbE product and related sales totalling approximately GBP8,000,000. LPA sold the technology to global companies Nomad Digital and Icomera Solutions, providing transport connectivity solutions for trains [text removed for publication]. This technology enables rail operators to add new communications systems to their trains, delivering significant improvements to passenger experience and safety through better WiFi connectivity, improved passenger information systems, HD CCTV and on-demand infotainment .

2. Underpinning research

By the early 2000s, rail journey times became competitive with short haul flights, however, Internet connectivity still lagged behind in the harsher train environment. Passenger experience, security improvements (e.g., CCTV) and improved on-train communications required increased capacity however, 100MbE download rates seemed unsurpassable. Research led by Walker at Essex, developed a rail approved 10GbE Ethernet communication system, now widely installed on trains, which exceeds the capacity the rail industry considered best possible by two orders of magnitude.

From research commencing in 2004, Walker presented [R1] an optically remoted leaky feeder system for data transmission in hostile environments like the underground transportation networks. The 1.6Gb/s data throughput, the highest reported in this application at that time, offered future-proof provision for broadband security networks. This research in data transmission and service provision in the hostile rail environment subsequently contributed to a solution for the radio frequency (RF) cross-talk electro-magnetic field, which affects the existing train connectors, as it provided the tools to study the problem. The measured 2.4 - 6 GHz leaky feeder bandwidth allowed multi-radio, multi-band wireless mesh networking and therefore low contention Gigabit rolling stock data communications. For the in-carriage environment, especially improving in-train data distribution, as is the case here, Walker’s research into full service access and edge networks [R2] demonstrated for the first time that Gigabit throughput using the emerging 5G 24-GHz frequency band is an option for future in-carriage wireless networks. Communications through copper media, widely used and considered reliable by the rail industry, and wireless transmission systems insights [R3] presented for the first time a novel implementation of 4K UHD live video encoding for streaming over a wireless network at low bitrate indoors, using GPUs for parallel H264/AVC video encoding, in order to enhance the on-train infotainment experience.

Progress towards fulfilling expectations for increased capacity using wireless, wired and hybrid communication systems in non-trivial real-world railway applications [R1] combined with [R2], [R3] provided the foundations for collaboration with LPA, with significant and challenging requirements of achieving 10GbE networking and managing associated issues, including cross-talk; RF data transmission (particularly 2.5 - 2.7GHz range) and realising opportunities and challenges of 4G and 5Gmobile networks; fibre-optics and inter-carriage wireless systems. Consequently, Walker collaborated with LPA [G1] to enable them to design rail connectors for RF communication and Ethernet data transmission. Most importantly, [G1] enabled Walker to develop the new Ethernet backbone. Walker addressed challenges of: 1) cabling and connectorisation at gigabit data rates; 2) cross-talk issues which meant ensuring inter-pair shielding for multi-pair cables, whilst 3) retaining compliance with standard connector sizes. Though practical investigation, a new system was developed to withstand the rugged train environment [R4, G1].

A primary issue for Ethernet cabling and connectors, as identified in [R1], is near-end crosstalk (NEXT), where cable pairs interact both capacitively and inductively to produce radiative interference [R1, R4], which degrades the data throughput, and deteriorates with increased data rate. The IEEE standard 10GBASE-T network interface card (NIC) deals with NEXT, inter alia, very effectively. Thus, NEXT is not an issue in data centres where standardised cable (CAT 7), plugs and sockets are used. In the hostile rail environment, however, severe cable flexing must be accommodated, e.g., for use in inter-carriage jumpers where CAT 7 cable fails mechanically. At the start the study [G1] in 2013, connectors, adequate for 100MbE Ethernet, were using industrial, super-flexible CAT 5 cabling. Essex’s investigations with FLUKE 100MbE to 10GbE certified test equipment (which extracts sophisticated diagnostic information) revealed CAT5 cable failed beyond 1GbE. Worst still, the existing connectors failed NEXT levels at just over 100MbE. The NEXT failings of existing cabling and connectors were solved separately. For the cables, Walker’s transmission line insights [R1] were applied. It was found that a short length of rugged, flexible, industrial CAT5 cable (for use in jumpers) could be attached to a much longer CAT7 cable (~ 25m, for use in carriages) to restore 10GbE cable performance.

The connector’s NEXT problem was ultimately solved by research and analysis of RF systems [R1, R4, R5]. This showed that a balanced quadrupole arrangement (instead of the previous unbalanced star-quad deployment, common in the rail industry) for the connecting pins in their housing was needed. This produced cancellation of both the electric and magnetic fields in the centre of the connector pins, which were arranged in groups of four on a square template. The key result was the breakthrough realization that zero electrical and magnetic field in the middle of the connector could not generate any cross talk in the surrounding cables [R4, R5]. Initial tests with the FLUKE apparatus showed full 10GbE capability for combined 25m CAT7 and 3m CAT5 entry and exit cables through one connector, with the required wiring modification. The tests were extended to the full 10GBASE-T 100M length (four cable spans and three connectorized jumpers), again with a satisfactory result (pass certificate). Industrial 10GbE switches were then utilised for an Ethernet Train backbone (ETB) test equivalent to 10 carriages. After testing with selected rail operators, the 10GbE ETB system was successfully demonstrated. [R5, G2] also demonstrates the 10GbE ETB enhanced to 40GBE (by spatial multiplexing), shown to work with data-centre calibre PCs.

3. References to the research

[can be supplied by HEI on request]

R1 S. E. M. Dudley, T. J. Quinlan, and S.D. Walker. “Ultra-broadband Wireless–Optical Transmission Links Using Axial Slot Leaky Feeders and Optical Fiber for Underground Transport Topologies” 2014 IEEE Transactions on Vehicular Technology, Volume: 57, Issue: 6, 2008, 3471 – 3476 DOI: 10.1109/TVT.2008.920055

R2 O. J. Femi-Jemilohun, T Quinlan, S. Barc, and S. D. Walker. “An Experimental Investigation into GbE Wireless Data Communication at 24GHz in Non-Line-of-Sight and Multi-Path Rich Environments.” Antennas and Wireless Propagation Letters: December 2014 Volume: 13 Issue: 1, 1219-1222. DOI: 10.1109/LAWP.2014.2332236

R3 O. Adeyemi-Ejeye and S. Walker “4K UHD H264 Wireless Live Video Streaming Using CUDA,” Journal of Electrical and Computer Engineering, vol. 2014, Article ID 183716, 12 pages. (Published February 2014). DOI: 10.1155/2014/183716

R4 G. Koczian, S. Walker, G. Howell, and B. Simpkin . “A 10 Gbit/s Ethernet Infrastructure for future-proofed railway communications.” The Stephenson Conference: Research for railway, 21 - 23 April 2015. Institution of Mechanical Engineers, London.

R5 F. Ngobigha, S. D. Walker, G. Koczian, G. Howell, and J. Prentice. “Demonstration of 40 Gbit/s conducting media data capacity on international rolling stock,” 15th Annual Conference on Wireless On-demand Network Systems and Services (WONS), January 2019. DOI: 10.23919/WONS.2019.8795504 http://dl.ifip.org/db/conf/wons/wons2019/115.pdf

G1 Walker, S. D., Woods, J.C. KTP8968 - To enable LPA to design rail connectors suitable for RF, LPA Industries Ltd and the Technology Strategy Board, 07 Jan 2013 - 06 Jan 2015, GBP40,613 and GBP82,456, GBP123,069.

G2 Walker, S. D., Woods, J.C. KTP010573 - To embed a novel wireless Ethernet capability across the rail network, to support innovation in train to shore communication speed and establish LPA as a leading innovator in the sector. To increase the bandwidth of on-board copper Ethernet Backbones, LPA Group PLC and the Technology Strategy Board, 30 Jan 2017 - 22 Jan 2020, GBP75,687 and GBP153,667, GBP229,354.

4. Details of the impact

Essex research enables company LPA commercial success to become a world leader for rail ethernet technology, achieving over 60% of the market share for advanced rail inter-car connection systems

LPA Connection Systems is a UK based company designing and manufacturing inter-connection systems and electrical control boxes for railway rolling stock. In 2013, the rail market was demanding higher data transmission rates for its Ethernet systems, capacity was previously only 100MbE, to permit faster internet access for passengers and to improve railway security by enabling HD CCTV images to be recorded centrally on the train. Essex research [R1-R5] showed how data centre copper cable technology can be applied successfully in the railway environment and via a collaborative development partnership with Essex funded by Innovate UK [G1], LPA wasable to offer the world's first 10GbE bandwidth Ethernet train backbone over copper’* [S1][S2]. LPA launched this first rolling-stock approved 10GbE technology in 2014 with an order to retrofit a large UK fleet of trains and now supplies its market leading 10GbE products to customers across the world [S3]. The COO of LPA states ‘ Being the first to introduce 10G technology into an intercar jumper enabled LPA to sell more systems than would normally be seen’ adding **‘ Harnessing expertise in the University of Essex, the KTP enabled LPA to move into a new technology area and develop a highly successful product that has sold and continues to sell worldwide ‘[S1].

In 2010, before the collaboration with Essex, LPA’s company turnover was GBP4-6 Million p.a., which increased to GBP6-9 Million p.a. (c60%) by 2017, ‘ largely as a result of its Ethernet technology products’, based on the work with Walker at Essex [S3]. Sales increased accordingly; LPA not only achieved direct sales of 10GbE Ethernet Inter-car Jumpers but also was able to leverage this unique 10GbE technology to sell complete LPA Inter-car connection systems to its customers [S3] and already by 2017 ‘ The research led by Prof Walker at the University of Essex […] provided an essential contribution to £8M worth of sales for LPA’** [S3, S4, S5]. The development of the 10GbE technology led to LPA increasing its market share in the intercar jumper market as stated in the final KTP report *‘ *The demonstration of 10GbE represented a significant step change in performance for the industry and catapulted LPA to world leader for Rail Ethernet technology’ [S2, p.2, p.4, p.8 and p.10] . [text removed for publication]

As a result of the Essex research, **LPA has enhanced its knowledge & capabilities in data communications which ‘ *enabled LPA to move into a new technology area’ [S1]. ‘ The company has acquired in depth knowledge in electronic testing processes and 10GbE networking and associated issues, as well as new knowledge in RF data transmission and mobile networks. The Company has acquired knowledge about the process of identifying new technologies and testing their feasibility’ [S2, p.5]. LPA established a new electronics development facility [S2, p.5] and through staff training ‘the technological know-how has been embedded within the company’ [S1].

LPA’s customers benefit from futureproofed 10GbE systems for trains enabling improvements to rail security and passenger experience

LPA’s 10GbE products were sold to a range of rail industry customers operating globally [S3]:

  • Systems integration companies such as Icomera, the world’s leading provider of wireless internet connectivity and application platforms for passenger transport and Nomad Digital which operates widely across Europe, North America and Australia [S6] [S3].

  • Rolling stock operators [text removed for publication] operating in 38 countries worldwide [S6]; [text removed for publication]

  • Rail operators: In the UK LPA 10GbE products are used by almost a third of the UK’s rail operators [S6]: [text removed for publication]

Between 2014 and 2019, over 1,000 train cars were fitted with 10GbE technology [S1]. By enabling LPA’s customers to purchase and install 10GbE-ready jumpers and fixed harnesses to their trains for the same price as a 100MbE system, they have been able to futureproof their rolling stock and avoid the large cost and fleet disruption associated with fitting upgraded jumpers when data demand on-board trains overloaded their 100MbE capability. ‘ These systems have enabled operators to futureproof their vehicles and ensure adequate data capability for any systems that may be added at a later point’ LPA COO [S1]. The Technical Manager of Angel Trains, one of Britain's leading rolling stock companies, which leases stock to 20 UK rail operators [S6] states ‘ With the development and launch of LPA’s 10GbE-Ready Jumpers, the justification for retrofitting of 10GbE-Ready Inter-car Jumpers was made easier’, adding ‘ With the future-proofed Ethernet Jumpers, further increases in data transmission demands in the future can be accommodated without another expensive retrofit programme being carried out on trains - both expensive (>>£1M) and necessitating trains to be taken out of service one-by-one*’ [S7].

Rail passengers increasingly require high-speed on-train connectivity that supports entertainment, social media engagement and business activity and identify WiFi as a reason for choosing a particular rail operator. As a result ‘ Train operators continue to invest in next generation hardware platforms and upgrades to the on-train network, including migration to a 10GbE backbone’ according to Nomad Digital’s Head of Product Management [S8]. This continued investment allows operators to benefit from services such as CCTV, passenger WiFi and sophisticated fleet management solutions and enables a new raft of entertainment services to be deployed [S8]. The Angel Trains Technical Manager confirms the benefits to the rail operators and their passengers: ‘ The LPA 10GbE-ready backbone enables our customers [rail operators] to exploit the future-proofed data capacity to enable new communications systems to be added to their trains to deliver significant improvements to passenger experience; for example, Wi-Fi, improved Passenger Information Systems, on-demand infotainment, passenger counting, digital HD/4K CCTV and Big Data analytics**’ [S7]. Rail operators and passengers also benefit from more reliable trains through the use of the 10GbE enabled technology ‘ Improved train reliability initiatives benefit from preventative maintenance regimes that exploit data from sensors all over the train, for instance to flag up a potential bearing failure before it occurs. Angel Trains’ Class 357/2 fleet was upgraded with McLaren’s “Intelligent Train” system, with the objective of using McLaren’s Big Data analytics to monitor the train sensors and report for any abnormal outputs. This 1GbE system communicates over an LPA 10GbE-Ready backbone’ [S7]. Angel Trains concludes that: ‘ In the rail industry, 10GbE was a leap in data transmission capability and the resulting bandwidth benefits that these technical innovations have since bought to Angel Trains, the Train Operating Companies and their passengers would otherwise not have been possible’ [S7].

5. Sources to corroborate the impact

[S1] Testimonial from the Chief Operating Officer, LPA Connection Systems

[S2] Final Report KTP: KTP8968

[S3] Testimonial from the MD, Rail innovations Ltd (former MD of LPA 2009-2019)

[S4] Article from Rail Engineer 2017

https://www.railengineer.co.uk/innovation-conference-gets-bigger-and-better/

[S5] E mail testimonial MD, Rail innovations Ltd (former MD of LPA 2009-2019)

[S6] Scale of Operations of LPA’s customers (complied pdf)

[S7] Testimonial from the Angel Trains Product Technology Manager.

[S8] Article, Head of Product Management Nomad Digital

https://nomad-digital.com/articles/passenger-expectations-in-todays-connected-world/

Submitting institution
The University of Essex
Unit of assessment
11 - Computer Science and Informatics
Summary impact type
Societal
Is this case study continued from a case study submitted in 2014?
No

1. Summary of the impact

Essex NLIP research resulted in the development of an online platform for secure real-time civilian-led reporting of human rights violations in conflict zones providing more accurate monitoring in areas inaccessible to traditional human rights monitoring missions. International human rights organisation, Minority Rights Group International (MRG) and the Ceasefire Centre for Civilian Rights, deployed the platform between 2017 and 2019 in Iraq, Iran, Syria, the Middle East and North Africa. In Iraq more than 3,700 cases of human rights violations were documented using the platform since 2017. The platform has enabled MRG to grow its activities and was essential for the establishment and operation of Ceasefire, furthering the practice of civilian-based monitoring of violations. This secure and easy to use platform has empowered and enabled ordinary civilians to document violations, mitigating the risks associated with reporting of human rights abuses in conflict situations. The evidence of human rights violations obtained from the platform has enabled documentation of claims for reparations, to compensate the victims of human rights abuses (cases documented on the platform can form the basis for reparations claims under the Iraqi national Law 20 reparations system, which has paid out over USD 355 million in compensation) and for international oversight, generating reports which have been used by the United Nations (UN), ultimately improving human rights outcomes for civilians.

2. Underpinning research

Essex Natural Language Information Processing (NLIP) researchers collaborated with MRG and the Ceasefire Centre for Civilian Rights (Ceasefire hereafter) from 2014 to design and implement a new platform to support civilian-led monitoring of human rights violations [G1]. Their challenge was to develop a reporting platform which would enable civilian-led reporting in inaccessible areas which would be secure, easy to use and be able to assess the accuracy of reports, supporting both English and Arabic languages. Developing a platform for this purpose was particularly challenging as Arabic is an under-resourced language in NLP and requires a unique set of skills. The research of the NLIP team, on information extraction from social media, Arabic natural language processing and crowdsourcing addressed these challenges.

The Essex research provided insights into crowdsourcing which focused on determining a "gold standard" based on a large number of individual statements with varying degree of reliability e.g., [R1]. The reporting platform was a development of Ushadhidi, an open-source crowdsourcing platform used to map reports of violence, but its security model was insufficient for Ceasefire's needs. The Essex researchers developed a new security model and user access control, providing additional structural support for collaborating organisations with additional structural and security modifications [R2]. Essex’s Arabic natural language processing toolkit (AraNLP) [R3] together with methodology to identify named entities (e.g., names of people and places) which describes how names, places and dates can be extracted [R4], provides the basis of the Arabic text processing pipeline that automatically processes social media texts to identify potential human rights violations. AI-powered approaches for classifying massive amounts of data into pre-defined categories need enough reliable training data and machine learning algorithms of sufficiently high classification quality. This was addressed by Essex research on extracting and managing content from social media, using crowdsourcing techniques to label and evaluate the collected data and the framework for using crowdsourcing to create language resources for under-resourced languages such as Arabic [R5] [R6]. The Essex language resource of Arabic texts representing different types of violence including human-rights violations [R7] was the first such resource. This addressed the lack of previous work on detecting human rights violation in English and Arabic, the modest accuracy of existing methods and the paucity in robust and scalable resources to process Arabic natural language. One reason for this is the difficulty in applying methods developed for languages such as English to Arabic, a highly derivational and inflectional language that is characterised by a complex set of morphological features including gender, number, person, case, state, mood, and voice. In addition, Arabic has a set of clitics, which attach to the stem after affixes such as conjunctions, prepositions, future marks, definite articles, and pronouns. [R2] reports Essex’s insights into mining and classifying Arabic Twitter in order to identify potential human rights abuse incidents in a continuous stream of social media data within a specified geographical region. Essex’s analysis showed deep learning approaches such as Long Short-Term Memory (LSTM) classification precision was 84% with an F1-score of 75%. The practical usefulness of the Twitter mining tool for the analysts’ work was also demonstrated and was deemed of high enough quality to be used in the practical setting [R2].

This Essex research resulted in a new platform for Ceasefire (Fig.1), described in [R2], which monitors human rights violations within a specified geographic area via (i) a structured reporting system which experts and individual witnesses submit to, enabling civilian-led monitoring, combined with (ii) AI mining of social media yielding a continuous stream of data which have been classified as signals of potential violations within the same region. Essex's system constituted the portal first portal of its kind which combines (i) and (ii). The platform’s sources include reports submitted by users directly to the website, email and SMS, Twitter, and mainstream news reports, in both Arabic and English [R2]. Reports submitted through the form are stripped of personally identifying information and plotted onto a live, interactive map visible on the tool’s landing page, which shows the distribution of violations by location and type. Sensitive personal data is stored securely on a secondary server in case it is needed for future follow-up or court proceedings. The reports obtained are made available to partner organisations and the public at the same portal web interface where reports are submitted.

Fig. 1: Ceasefire: a framework for reporting and monitoring of human rights abuses.

Embedded image

3. References to the research

[can be supplied by HEI on request]

This research was published in respectable books, journals and conferences with global reach. All papers were peer-reviewed and are internationally recognised for the originality, significance and rigour of the research.

[ R1] T. Fornaciari and M. Poesio. 2014.Identifying fake Amazon reviews as learning from crowds. In Proceedings of EACL, Gothenburg, April 26-30 2014. https://dx.doi.org/10.3115/v1/E14-1030

[ R2] A. Alhelbawy, M. Lattimer, U. Kruschwitz, C. Fox, and M. Poesio. 2020. An NLP-Powered Human Rights Monitoring Platform. Expert Systems with Applications, Volume 153, 113365. https://doi.org/10.1016/j.eswa.2020.113365.

[ R3] M. Althobaiti, U. Kruschwitz and M. Poesio. 2014. AraNLP: a Java-based Library for the Processing of Arabic Text. In Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC), Reykjavik, Iceland, 2014. http://www.lrec-conf.org/proceedings/lrec2014/pdf/621_Paper.pdf

[ R4] M. Althobaiti, U. Kruschwitz, and M. Poesio. 2015. Combining Minimally-supervised Methods for Arabic Named Entity Recognition. Transactions of the ACL (TACL), 3:243–255, 2015. https://doi.org/10.1162/tacl_a_00136

[ R5] M. Poesio, J. Chamberlain and U. Kruschwitz. Crowdsourcing. In N. Ide and J. Pustejovsky (eds.), Handbook of Linguistic Annotation (Springer), 2015. ISBN 978-94-024-0881-2

[ R6] M. El-Haj, U. Kruschwitz, and C. Fox. 2015. Creating language resources for under-resourced languages: methodologies, and experiments with Arabic. Language Resources and Evaluation. 49 (3), 549-580. https://doi.org/10.1007/s10579-014-9274-3

[ R7] A. Alhelbawy, M. Poesio, and U. Kruschwitz. 2016. Towards a Corpus of Violence Acts in Arabic Social Media. In Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC), Portoroz, Slovenia, 2016. http://www.lrec-conf.org/proceedings/lrec2016/pdf/553_Paper.pdf

Grants:

[ G1] KTP InnovateUK grant KTP009488 with Minority Rights Group International (MRG). “To Design & Implement New Systems to Support Civilian-Led monitoring of human rights violations and embed the capability to enhance and exploit the system in the future.” Udo Kruschwitz (PI), Massimo Poesio (CoI), October 2014 to September 2017, £210,320.

[ G2] McGregor, Sunkin, Fussey, Poesio, Bhalotra, Leader, McDonald-Maier, Human Rights and Information Technology in an Era of Big Data (The Human Rights, Big Data and Technology Project), ESRC, October 2015 – September 2021, £4,743,734 with additional funding from the University of Essex

4. Details of the impact

Developing and implementing a secure reporting platform enabling civilians to document human rights violations

The ability of traditional human rights monitoring via investigative missions to work effectively and achieve their aims is challenged by restricted access to insecure territories, as well as cost, time, questions over politicisation, oversight, legitimacy and representation. With advances in communications technology, civilian-led monitoring in conflict zones has become a complementary, and sometimes the principal, source of information on human rights violations [S1]. Although new technologies provided hardware for civilians to report violations in theory, obtaining reliable human rights data and information that can be documented required a new system. Essex NLIP researchers collaborated with MRG (a leading international human rights organisation, with over 150 partners across 60 countries) [S2] and Ceasefire (founded in 2014 as an in-house centre of specialist expertise on civilian-led monitoring within MRG) on the 2014-2017 Ceasefire Programme to develop and implement such a system of civilian-led monitoring of human rights abuses [S3].

The Executive Director (ED) of Ceasefire (and former ED of MRG until 2018) confirms that ‘ Essex’s NLIP research was applied to develop the essential technological infrastructure to support civilian-led monitoring, including dual Arabic/English language reporting, as well as techniques for social media data mining*. Specifically, they developed an integrated online platform to receive and process information from sources, including reports submitted by users directly to the website, email and SMS, information posted on Twitter, and mainstream news reports, in both Arabic and English which was critical to the Ceasefire programme’s success’ [R2] [S3]. They add: ‘ Essex's specialist insights in Arabic NLP, working in areas such as text summarisation and named entity extraction, provided the essential foundations for developing the platform’ [e.g. R3] and ‘ through further research in a project funded by MRG and Innovate UK, the NLIP researchers constructed the online Ceasefire platform […] Essex’s NLIP research was the essential basis for that project’s results’ [R1-7] [S3]. The Ceasefire platform comprised two technological developments: A violations reporting form, allowing users to submit reports of violations directly to the website and the addition of social media data mining tools to the website, allowing information on human rights violations to be automatically harvested from Arabic language Twitter feeds [R2] [S3]. Importantly ‘ Essex’s NLIP research provided the capability to adapt and enhance the systems for new environments, and in line with changing modes of communication*’ so that the reporting platform is ‘ applicable to different contexts, and conflict situations’ [S3]. The development of this new platform wasessential for the success of the Ceasefire programme which has empowered civilians subject to violations in highly insecure environments’** [S3].

This reporting platform was piloted by MRG and Ceasefire in Iraq in 2016 with MRG’s partner organisations, including Asuda - Combating Violence against Women, then publicly rolled out in Iraq in 2017, Iran in 2018, part of Syria in 2019 and more widely across the Middle East and North Africa in 2019 [S3]. Accessible to anyone with an internet connection, the platform empowers and enables ordinary civilians to document violations of their rights in a way that conforms to international legal standards and increases the future usability of the data [S3, S4 p.57]. The creation of a safe, secure and easy to use platform [R2] has helped to mitigate the risks to those reporting human rights abuses in conflict situations, including civilians, journalists and humanitarian agencies [S4]. ‘ *In the current situation in Iraq, the major risk facing human rights activists and defenders is reporting violations [...] The online tool helps them prevent these risks as they easily use the tool and it is anonymous.*’ Director of an Iraqi NGO [S4 p 58]. For civilians, the platform enables them to report incidents at any time and in a more confidential way than talking with a representative of an NGO, which could be under surveillance [S4]. In Iraq more than 3,700 cases of human rights violations were documented using the platform since 2017 [S3, S4 p. 58], several hundred in Syria since 2019 – ‘ a considerable body of documentary evidence of violations’** [S3]. In addition to the violations reported via the platform, information harvested from social media alerts the relevant authorities to violations, providing a means of early warning [S3]. ‘ The Twitter crawler also provides real-time information on the popular reaction to reports of violations, giving us a perspective on people’s attitudes to human rights abuses, including at local level’ [S3].

Contributing to the work of UN bodies, increasing the availability of reparations and improving human rights outcomes

Data obtained from the reporting platform contributes evidence for influential reports [e.g. S5a, b] submitted by MRG/Ceasefire to UN bodies for international oversight of human rights issues. Ceasefire’s ED states ‘ Different UN mechanisms use the information provided as a result of the tool and some have referred to our reports which the platform provided essential evidence for*’ [S3]. For example, ‘our work on Iran, based partly on one of the tools, was quoted by the UN Secretary-General in his report on Iran for the UN General Assembly in 2019 [see S6 p15] . This is the most significant and authoritative annual report on human rights in the country’ [S3]. When protests erupted in Iraq in 2018, reports of excessive use of force by security forces against peaceful protesters began to appear on platform. These cases were incorporated into a wider report on repression of civilian activists in Iraq [S5a] which was then used as a basis for urgent appeals and advocacy pushes at the UN Human Rights Council and other forums [S4]. Ceasefire have a strong record of having their recommendations on human rights adopted: ‘we made a total of 63 separate recommendations on Iraq to the UN Universal Periodic Review, the UN Committee on the Elimination of Racial Discrimination and the UN Committee on the Elimination of Discrimination against Women; of these, 28 were fully adopted and 16 partially adopted by the relevant UN body’ Ceasefire ED [S3].

The evidence of human rights violations obtained from the platform has also enabled documentation of claims for reparations, to compensate the victims of human rights abuses [S3]. ‘ Cases documented on the platform can form the basis for reparations claims under the Iraqi national Law 20 reparations system, which has paid out over USD 355 million in compensation. This aided our cooperation with the UN assistance mission in Iraq and with the Iraqi Parliament: we subsequently formed an agreement with the Iraqi Human Rights Committee to provide technical assistance on the drafting of new human rights legislation. We also have a developing relationship with IOM (International Organization for Migration) because they have taken a particular interest in reparations.’* Ceasefire ED [S3]. By enabling the development of the platform, the Essex NLIP research has ultimately ‘ empowered civilians subject to violations in highly insecure environments and boosted the availability of reparations. This had a major effect on human rights outcomes’** Ceasefire ED [S3].

Supporting MRG’s human rights operations and the establishment of Ceasefire furthering the practice of civil society based monitoring of violations

Through the development of the new platform, Essex research enabled MRG **‘ to develop a major new area of expertise’* [S7] and ‘provided the technology for the system to enable civilian-led reporting of human rights abuses which Ceasefire’s work depends upon and was thereby an essential factor in establishing and maintaining Ceasefire, furthering the practice of civil society based monitoring of violations. This in turn helped Ceasefire and MRG to secure significant new external funds for related projects, including the award of a new EUR1,000,000 project by the EU’ [S3], [S7, p4]. MRG is financed by donations and grants from foreign governments, trusts and foundations, and from individual donations [S7, p2]. The increase in its income from a year before the project began to 2017 was almost entirely attributed to the platform (S7, p5-6, items 9 and 10, pages 5-6). MRG’s staff acquired new skills in the management of online reporting platforms, improved understanding of user needs and behaviour in the reporting of human rights violations on the ground, better appreciation of the adaptation of international standards on human rights monitoring and reporting, as well as a wider awareness of digital advances to support the human rights movement generally [S7 p5]. Furthermore, ‘ training in the use of the reporting platform has been undertaken on two occasions for civil society organisations in Iraq’ [S7 p5]. MRG also employed a new Programme coordinator to implement the results of the project [S7 p8].

5. Sources to corroborate the impact

[S1] Eyes on the Ground: Realizing the potential of civilian-led monitoring in armed conflict (2017)

[S2] Minority Rights Group Website https://minorityrights.org/about-us/

[S3] Testimonial from Executive Director (and former Executive Director of MRG), Ceasefire

[S4] 50 Years of Minority Rights Group Report (2020)

[S5] a. ‘Civilian Activists under Threat in Iraq’ report (2018) b. ‘Cultivating Chaos: Afrin after Operation Olive Branch’ report (2020)

[S6] UN General Assembly, Report on Human Rights in the Islamic Republic of Iran, 8 February 2019, A/HRC/40/24

[S7] Final Report on Knowledge Transfer Partnership KTP009488

Submitting institution
The University of Essex
Unit of assessment
11 - Computer Science and Informatics
Summary impact type
Technological
Is this case study continued from a case study submitted in 2014?
No

1. Summary of the impact

Essex research led by Hagras improved the maintenance efficiency [text removed for publication]. The unique iPatch tool designs [text removed for publication] BT [text removed for publication] engineers’ working areas to maximise productivity, minimise travel and best match skills to tasks. This increased engineers’ productivity by 7%, reduced fuel consumption by [text removed for publication] 2000 metric tons of CO2 emissions [text removed for publication]. The work also contributed to [text removed for publication] Parliamentary debates on AI.

2. Underpinning research

Optimal deployment of engineers to maximise the number of tasks the workforce completes is essential for successful maintenance and development of communications infrastructure. Consequently, engineers’ work areas (patches) are designed to maximise the proportion of their time spent completing tasks. This means minimising time spent travelling between tasks by reducing distances (route planner road travel times) between them, whilst minimising variation between the work hours patches contain.

Real world decisions are often made on imprecise information. Fuzzy Logic extends Boolean Logic to handle partial truth with truth values being a real number between 0 and 1 inclusive, ranging from completely true to completely false. Fuzzy models or sets enable better handling of imprecise information and uncertainty. Type-1 FLSs (Fuzzy Logic Systems), applied to many real-world problems cannot fully handle the numerous uncertainties encountered in real world environments. Higher order fuzzy logic systems such as general type-2 FLSs, have been shown to be eminently suited to deal with such high levels of linguistic and numerical uncertainties but until recently the great computational complexities associated with general type-2 FLSs prevented their application to real-world problems. In addition, real-world workforce optimisation needs efficient and robust many-objective optimisation systems which can converge rapidly to satisfy large sets of conflicting objectives and constraints.

In 2010, Hagras introduced a pioneering theoretical approach: a complete representation framework, referred to as zSlices based general type-2 fuzzy systems [R1]. The paper was awarded by the 2010 outstanding paper in the IEEE Transactions on Fuzzy Systems. As its award citation states: “ This seminal paper allowed breakthroughs to the theory and applications of type-2 FLSs where the paper presented a complete “modern” approach to design and realize general type-2 FLSs based on zSlices type-2 fuzzy sets”. This significantly reduced both the complexity and the computational requirements for general type-2 FLSs and provided the capability to represent complex general type-2 fuzzy sets, thus allowing, for the first time, the realization of general type-2 FLSs for real world applications. This resulted in a revolutionary new wave of FLSs that can handle the high levels of uncertainty present in real world applications - something which is not possible by other means. This powerful property, in combination with their implementation and computational simplicity, yielded a powerful new generation of intelligent systems which outperform their type-1 and interval type-2 FLSs counterparts in performance.

BT’s spreadsheet and map-based design of working areas (patches) for deploying engineers was extremely time consuming (in the order of months) and resulting in non-optimal results, so in 2011, Hagras began research funded by BT to optimise the strategic deployment of its engineers for better service delivery. Moreover, Hagras’ unique research in Multi [R2], [P1] and Many Objective Type-2 Fuzzy Logic Based System for workforce allocation presented a novel component termed Fuzzy Dominance Rules [R3], [P2], [P3] which addressed the weaknesses in standard multi-objective optimisation systems. These systems allow the design to be “evolved” by evaluating millions of different possibilities and progressively moving towards better design. Meanwhile, moving engineers between teams to balance each team’s skills pool was addressed in [R4]. The combination of Hagras’ research in general type-2 FLS and many objective optimisation led to novel systems, which better handle the uncertainties and conflicting multiple objectives present in real-world workforce allocation problems to deliver a system capable of robustly satisfying the conflicting objectives in workforce optimisation.

3. References to the research

[can be supplied by HEI on request]

[R1] C. Wagner and H. Hagras (2010) Towards general type-2 fuzzy logic systems based on zSlices, IEEE Transactions on Fuzzy Systems, 18, 637-660. (375 citations – February 2021).

https://doi.org/10.1109/TFUZZ.2010.2045386 This paper was awarded, by IEEE Computational Intelligence Society (CIS), the 2010 IEEE Transactions on Fuzzy Systems Outstanding Paper Award.

[R2] A. Starkey, H. Hagras, S. Shakya, G. Owusu (2016) “A Multi-Objective Genetic Type-2 Fuzzy Logic Based System for Mobile Field Workforce Area Optimization” (60 citations – February 2020) Journal of Information Sciences, Vol. 333, pp. 390-411. https://doi.org/10.1016/j.ins.2015.09.014

[R3] A. Starkey, H. Hagras, S.Shakya, G. Owusu (2019) “iPatch: A Many-Objective Type-2 Fuzzy Logic System for Field Workforce Optimisation,” (9 citations – February 2021) IEEE Transactions on Fuzzy Systems, Vol. 27, No.3. https://doi.org/10.1109/TFUZZ.2018.2862394.

[R4] A. Starkey, H. Hagras, S. Shakya, G. Owusu, (2016) A Genetic Algorithm Based Approach for the Simultaneous Optimisation of Workforce Skill Sets and Team Allocation. In: Bramer M., Petridis M. (eds) Research and Development in Intelligent Systems XXXIII. SGAI 2016. Springer, Cham. https://doi.org/10.1007/978-3-319-47175-4_19. Won the Best Paper Award in the Conference.

Patents

[P1] “Scheduling tasks to resources for a network using fuzzy logic”, Granted Patent US 10261833

[P2] “Optimisation of Delivery series over Communications Network”, Patent No, PCT/EP 3152659

[P3] “Method and Apparatus for retrieving a data package” Patent No PCT/ EP2018/061598

[text removed for publication]

4. Details of the impact

Workforce allocation tools developed at Essex with BT raised productivity by 7%, whilst reducing travel by 3%. Application of these tools within BT resulted in improved service to the vast majority of all UK digital network users.

PI Hagras and researchers at Essex developed the iPatch tool by applying Hagras’ breakthroughs in type-2 FLSs [R1], to create a system which designs optimal work patches to minimise travel [R2] and better match skills to tasks [R4], with novel new Fuzzy Dominance Rules [R3]. Working in collaboration with BT's researchers, Hagras’ research was applied to increase engineers' productivity by designing optimal work patches before publication. Practical implementation of this required close cooperation between Essex's researchers and BT's engineers, for instance the new working methods such as crossing county boundaries and change of culture, required approval of unions. iPatch soon improved performance [text removed for publication]. In a 2016 report on the iPatch [S2], BT’s Change Architect explained “ It’s one of a range of initiatives that’s contributed to better service performance including our best PSTN [public switched telephone network] performance in five years,".*

[text removed for publication]

BT does not just operate its own telecoms operation; it also operates the national network infrastructure via its network arm Openreach whose field force engineers maintain the copper wires that connect homes and businesses to phone and broadband. Openreach's customers include the 650+ communication service providers, which sell phone and broadband services to households and businesses [S3, p.2]. [text removed for publication] BT’s subsidiary EE Limited, services approximately 32 Million connections across its mobile, fixed and wholesale communications service networks and runs the UK's biggest and fastest mobile network. EE’s 4G coverage reaches 90% of the UK geography and 99% of the population. Their superfast fibre broadband service covers around 80% of the UK population, and their unlimited data allowance (ADSL) broadband service covers 98.7% of the population. As such EE is the largest and most advanced mobile digital communications company in Britain [S3, p.12-13]. Together with EE BT is one of the world's leading communications providers with for example 93% 4G coverage of the UK population [S3, p20]. [text removed for publication]

The importance of Essex's work on the iPatch was such that it was a major contribution to BT's Technology, Service & Operations (TSO) being named IT Project Team of the Year at the 2016 UK IT Industry Awards, one of the IT industry’s most coveted awards [S4]. Moreover, in addition to the benefits of optimising deployment and reducing travel time already mentioned BT notes [S4] that there was no solution on the market, which took an all-round approach to field analytics. [text removed for publication] Due to the innovative work with Essex, BT also won the IEEE Computational Intelligence society 2017 outstanding Organisation [S5]. Essex and BT’s collaboration won the 2015 and 2017 Global Telecoms Business Innovation (Business Service Innovation) Awards [S6]. [text removed for publication]

The iPatch research also contributed to the Scottish Parliament debates on how Artificial Intelligence can lead to future prosperity and contribute to UK growth in productivity and GDP. For instance, in the members’ business debate on motion S5M-10161 on Artificial Intelligence on 18th April 2018. In particular, the SNP's Kenneth Gibson, whose name the debate was in, opens by thanking the head of public affairs at BT and his colleague Dr Andrew Starkey 'for their excellent briefings' [S10].

[text removed for publication]

5. Sources to corroborate the impact

[S1] [text removed for publication]

[S2] Case Study by BT, Page 26, State of the Relationship Report 2016, National Centre for Universities and Business. Available at: https://www.ncub.co.uk/reports/state-of-the-relationship-report-2016.html

[S3] https://www.openreach.com/about-us/who-we-are [Accessed: 11.03.21]; https://ee.co.uk/our-company/about-ee [Accessed: 12/06/2020], Insight into TSO, BT.

[S4] BT Today, 2016. TSO wins award for best IT project team [Accessed 20 June 2019]

[S5] https://cis.ieee.org/getting-involved/awards/past-recipients#OutstandingOrganizationAward

[S6] Global Telecoms Awards: https://www1.essex.ac.uk/news/event.aspx?e_id=7695 and https://www.essex.ac.uk/news/2017/06/07/bt-collaboration-wins-global-telecoms-award

[S7] [text removed for publication]

[S8] [text removed for publication]

[S9] [text removed for publication]

[S10] AI Debate, Scottish Parliament April 2018: https://www.theyworkforyou.com/sp/?id=2018-04-18.20.2

Submitting institution
The University of Essex
Unit of assessment
11 - Computer Science and Informatics
Summary impact type
Technological
Is this case study continued from a case study submitted in 2014?
No

1. Summary of the impact

Essex research provided the crucial insights to create a novel media monitoring platform to monitor global news for business intelligence. The platform produced through research with global executive search and selection firm MBS Group, enables scalable automation of tailored news provision as well as delivering the high quality business intelligence companies demand, more widely. The platform proved so marketable that it enabled a new company, Signal AI to establish its commercial viability with rapid growth supported by Essex research in a collaboration, winning two national KTP awards. [text removed for publication]. Signal AI has since grown from 3 to over 150 full-time employees, [text removed for publication].

2. Underpinning research

Research led by Kruschwitz in the University of Essex's Natural Language and Information Processing (NLIP) research group demonstrated the capability for industry strength application of state of the art insights/research in Natural Language Processing (NLP) and Information Retrieval (IR). Specifically, research presented in [G1] led to the development of an enterprise-level intranet search engine for real world scalable applications [R1] and provided insights into improving suggestions to guide users [R2]. This demonstrated the capability required for [G1] through which a new media monitoring platform [R3] was developed. Subsequent research [R5, G3] advanced it further.

Work on supporting field operatives for British Telecom with contextual search [R1] progressed enterprise search level towards performance quality comparable to mainstream Internet search engines, i.e. those able to process large quantities of information in structured mark-up (typically HTML, but also PDF and Microsoft Word formats). The work in [R1] also provided insights into processing and indexing and then retrieving document collections residing in different data silos, which importantly, is analogous to processing a heterogeneous stream of news articles. In practice, the search engine in [R1] was developed and deployed for technical support staff. This ran live for several years and was accessed by the field force of the large telecommunications company. Kruschwitz et al.’s interactive adaptive query suggestion mechanism proposed [R1] guided users' searches through suggestions (for words and phrases users enter into a search box to obtain a list of results) derived from previous interactions. Further insights into generating such suggestions to enhance search, not of the whole web but smaller and more focused collections [R2] demonstrated: (1) the usefulness of log analysis to extract query modification suggestions; (2) a more fine-grained approach than grouping search requests into sessions allows for extraction of better refinement terms from query log files [R2].

Search method insights [R1, R2] resulted in [G2] by evidencing industry applicability. Research by Kruschwitz, Poesio and Martinez-Alvarez to build a scalable architecture for automated information provision resulted in a novel information search and filtering platform [R3]. This aggregates, analyses and classifies news articles so that they can be matched against a client's bespoke search profile. The cloud based architecture formed by an analytics pipeline comprising: Document Stream > Summarisation > Named Entity Recognition ( NER)/ entity recognition and disambiguation (ERD) > Topic Classification > De-duplication > Clustering > Email Delivery to Client. Each one of the 5 text analytics modules between document stream and Email delivery has a different component. Each document is processed through all the components, extending the information available for it. For instance, after the summarisation component, the system has access to the summary of the document. Near-duplication detection addresses the problem caused by republished articles in written media. Moreover, if not considered duplicates, often there are tens or hundreds of articles focused on the same information. These are addressed using clustering and event detection mechanisms. The pipeline uses a queuing system between components, allowing them to scale independently. This characteristic provides a scalable solution while minimising the complexity of the architecture. This also allows focus on specific solutions for each one of the components in order to improve the quality of the system over time [R3]. This addressed information filtering systems' growing problem of information overload, in particular to obtain insights to inform strategic decision making from disparate sources. The media monitoring platform [R3] allowed information search, filtering and summarisation on a scale and at a speed that was previously not possible as a single business service to analyse all the news of the world in real-time.

Essex NLIP researchers addressed the media monitoring platform’s next stage in development in [G3]. The new approach for scalable visualisation of sentiment and stance, which addressed scalable visualisation of planning data in e-government, [R5] was presented to Signal AI, demonstrating capability to address the problem in [G3]. Moreover, the first published paper into the problem of identifying whether a news article can be identified as topical or an aggregation of different news stories [R5], showed that the process can be broken down into a two-stage approach: first segment an article into smaller units, examine possible topic shifts and then apply a neural-network-based approach akin to image processing to identify whether the article/image resembles an aggregate or a topical piece. This advanced topic detection and filtering summary articles – key to the media monitoring system’s analytics modules. Narrow searches may exclude important information. Broader scope was exceeding existing monitoring products. Some articles, e.g., summaries should be excluded [R2]. Further work by Chamberlain and Brill on [G3] addressed turning a large feed of news articles into a digest of a small number of articles summarising and covering the key insights in those articles. They developed IR algorithms (based on clustering and ranking) for this (as noted in [S5]). [text removed for publication].

3. References to the research

[can be supplied by the HEI on request]

Following peer-review: [R1] was published by Springer and has been downloaded over 11,000 times; [R2] was published in a leading international information science journal (Q1, H124, IF2.7); [R3] (demo) and [R5] (full paper) were accepted at the annual European Conference on Information Retrieval (ECIR), the premier European forum for new research in the field of Information Retrieval. [R4] was presented at LREC, an important conference in the NLP community.

[R1] M-D. Albakour, G. Ducatel, and U. Kruschwitz. The Role of Search for Field Force Knowledge Management. In Transforming Field and Service Operations: Methodologies for Successful Technology-Driven Business Transformation, Theory and Applications of Natural Language Processing, p117–132. Springer, 2013. DOI: 10.1007/978-3-642-44970-3_8

.

[R2] U. Kruschwitz, D. Lungley, M-D. Albakour, and D. Song. (2013) Deriving query suggestions for site search. JASIST, 64: p1975-1994. DOI:10.1002/asi.22901.

[R3] M Martinez-Alvarez, U. Kruschwitz, W. Hall and M. Poesio. (2015) Signal: Advanced Real-Time Information Filtering. In: Hanbury A., Kazai G., Rauber A., Fuhr N. (eds) Advances in Information Retrieval. ECIR 2015. Lecture Notes in Computer Science, vol 9022. Springer, Cham. https://doi.org/10.1007/978-3-319-16354-3_87

[R4] J. Chamberlain, U. Kruschwitz & O. Hoeber. (2018) Scalable Visualisation of Sentiment and Stance. Proc. of LREC 2018, Myazaki, Japan. https://www.aclweb.org/anthology/L18-1660.pdf

[R5] M. Fisher, M-D. Albakour, U. Kruschwitz, and M. Martinez (2019). Recognising summary articles. In Proceedings of ECIR’19, 2019. DOI: 10.1007/978-3-030-15712-8_5.

[G1] EPSRC Grant. “Automatic Adaptation of Knowledge Structures for Assisted Information Seeking (AutoAdapt)”. Udo Kruschwitz (PI), 2008 to 2012, £278,271.

[G2] KTP InnovateUK Grant with MBS (Moira Benigson Executive Search LLP). “To build a scalable technology architecture that will enable automated information provision and creates a new revenue stream.” Kruschwitz (PI), Poesio (CoI), 2013 to 2015, £120,566.

[G3] KTP InnovateUK Grant with Signal Media. “Develop insight extraction and visualisation techniques to convert a stream of unstructured textual documents.” Kruschwitz (PI), Chamberlain (CoI) 2019 to 2020, £127,675.

4. Details of the impact

Development of a novel media monitoring platform contributed to business success for MBS and the rapid expansion of Signal AI

Company executives require regular, up-to-date news tailored to their business or sector to inform decision making. MBS Group (MBS), a leading global executive search and selection firm, found its manually produced newsletter was not providing the width and depth of coverage or the high quality, relevant, personalised external information required by its senior executives. [text removed for publication] Essex’s NLIP researchers demonstrated their research was applicable to this [R1, R2, G1]. Consequently, [text removed for publication] they began a project [G2] in 2013 to build the architecture required. Their resulting media monitoring platform is a state-of-the-art, cloud-based information processing architecture solution [R3] that not only enables scalable automation of providing tailored news but proved capable of delivering the high quality business intelligence companies demand more widely. The platform analyses world news, filtering out noise and focusing on trends and strategic information, users are provided with a structured stream of documents relevant to their business needs. [text removed for publication]

Importantly, through ongoing collaboration with Essex's NLIP researchers, the underlying technology for a new company, Signal AI, was established [text removed for publication]. According to Signal AI, the collaboration with the University of Essex ‘ created tangible value for Signal AI by allowing us to bridge the academic and industrial ecosystems in a seamless way through application of the University of Essex’s research. The first project allowed us to set the foundations of our Natural Language Processing (NLP) pipeline, Signal AI’s platform, by working in collaboration with Prof Udo Kruschwitz and Prof Massimo Poesio to apply their NLP expertise' [S2]. The project transformed the business so significantly and led to winning in 2015 the best Knowledge Transfer Partnership award of the year [S3]. Essex research was further embedded in Signal AI through Albakour becoming Data Scientist (in 2015) and Kruschwitz continuing as Signal AI’s Senior Advisor to ‘ advise and guide Signal AI in all research matters’ [S4].

Essex’s research was an essential part of Signal’s rapid expansion. Signal AI generated its first substantial sales to several well-known brands and organisations [text removed for publication]. This project ‘helped the company secure investment of over £5.8m. This played a pivotal and transformative role in the development of Signal’s AI-powered media monitoring capabilities as well as the company’s growth’ [S4]. In the second Innovate UK project with the company (started January 2019), Signal AI worked with Kruschwitz (until June 2019) and, leading thereafter, Chamberlain ‘who provided his expertise and advice, based on his research at Essex, not only on NLP but also on Human Computer Interaction (HCI) in order to maximise the value of the new developments in our product. This project started a new product line in the company' [S2]. Signal AI adds ‘ the collaboration provided us the support to move one level higher in the value chain, looking at how to provide more strategic insights to our clients based on our data’. [text removed for publication]

Commenting on the company’s growth, Signal AI note ‘During the course of these collaborations, Signal AI has gone from three people working in a garage to over 150 full-time employees now based in offices in London, New York and Hong Kong’ [S2]. They add that Signal AI’s collaborations with the University of Essex, as well as the features delivered as a result, have been a key selling point in discussions with investors ‘Signal AI has obtained more than £50,000,000 of investment to date, with the most recent Series C funding round of almost £20,000,000 completed in 2019’ [S2]. [text removed for publication] Signal AI developed a new feature, Signal Briefings, which provides clients with an executive level summary of a large feed of news to save time and keep clients updated with industry news [S5, S6]. [text removed for publication]. In reporting the next generation of AI technology in the Signal AI platform in February 2020, [S6] highlights ‘ Signal AI, one of the leading companies transforming how business leaders make sense of the world’s information' in 2019 'saw unprecedented success, more than doubling its customer base, international expansion into the US and Asia Pacific, and a successful Series C funding’. Signal conclude that ‘ Working with Essex's natural language processing and text analytics researchers gave our start-up the technical expertise needed to differentiate ourselves in a competitive market' CEO, Signal [S2]. This collaboration ‘ helped us drive innovation in the company’ [S2].

Signal AI’s [text removed for publication] global clients benefit from the Signal media-monitoring platform

[text removed for publication] The Signal AI platform has gone on to change the practice of its business clients in areas such as, tracking press coverage, monitoring global news, monitoring policy change and content generation and business intelligence [S7]. Public relations teams use Signal AI’s platform to track news coverage of their clients, for both promotional campaigns and to monitor negative news stories for reputation management and competitor performance. The platform has enabled MBS themselves, as well as leading international law firms such as Bird and Bird and Simmons and Simmons, to save time enabling them to process more information and adding value to their business. Bird & Bird’s PR team saved over an hour a day using the platform and note “ *We have halved our annual expenditure on media monitoring, but expanded and improved the speed of our coverage tracking.*” Head of Marketing and Communications, Bird & Bird [S7]. Simmons and Simmons, save 4-5 hours a month tracking news coverage, improving efficiency and enabling more information to be processed. The platform is also used to monitor enterprise-level legal policy and law. Deloitte, the largest professional services firm in the world, report that the platform enables over 1,200 of its clients to stay informed about changes proposed and enacted in law and regulation transforming the way they monitor and assess changes to tax law and regulation [S7].

Beyond business, the platform is used by NGOs such as Scholars At Risk, Amnesty International and the Disasters Emergency Committee to monitor and verify reports of human rights violations, and to deliver efficient responses to global disasters [S7]. Scholars At Risk, an organisation, which aims to protect threatened academic scholars around the world, report a 50% reduction in time spent monitoring global news for human rights violations, potential attacks on higher education institutes and to strategically plan campaigns and advocacy. The product has measurably helped their cause ‘ time saved is people saved’ Senior Program Officer, Scholars At Risk [S7]. Amnesty International is using the platform to verify eyewitness reports of human rights violations happening globally. Within the first few weeks of using this, they were able to quickly corroborate an eyewitness video of an extrajudicial execution in Mexico by locating a local Spanish language news story. Amnesty International was then able to compile sufficient evidence to call upon the authorities to perform an independent investigation. ‘ We were able to discover relevant news from sources we didn’t know existed’ Head of Digital Verification Corps at Amnesty International [S7]. The Signal product, enables the Disasters Emergency Committee, which coordinates the delivery of efficient disaster responses in the world’s poorest countries ‘ to have unlimited clippings during busy times without extra cost […] the flexibility of controlling our media monitoring service from our desktops allows valued agility to keep up with unpredictable and fast moving crises’ CEO, Disaster Emergency Committee [S7].

In conclusion, the CEO of Signal AI notes ' Our partnership with the University of Essex has been long and fruitful, and it has helped shape Signal into the innovative media company it is today.’ CEO, Signal AI [S2].

5. Sources to corroborate the impact

[S1] [text removed for publication]

[S2] Testimonial from Chief Data Scientist and quotes from CEO, Signal AI

[S3] Best KTP Award 2015

[S4] Signal AI Webpages Compilation

[S5] [text removed for publication]

[S6] Signal AI Introduces the Next Generation of Artificial Intelligence With the Launch of AIQ: https://www.businesswire.com/news/home/20200219005066/en/Signal-AI-Introduces-the-Next-Generation-of-Artificial-Intelligence-With-the-Launch-of-AIQ

[S7] Compilation of client case studies from Signal AI website:

https://www.signal-ai.com/customer-stories

Submitting institution
The University of Essex
Unit of assessment
11 - Computer Science and Informatics
Summary impact type
Technological
Is this case study continued from a case study submitted in 2014?
No

1. Summary of the impact

Essex researchers have pioneered new techniques for measuring the quality of speech and video transmitted over the Internet. A key contribution to the quality measurement was a new time-alignment method developed at Essex, in collaboration with Psytechnics. This work was incorporated into Microsoft’s product Skype for Business and by 2016 was used to measure the voice quality of approximately 100 Million users globally, including the US and UK, thus, improving the quality of teleconferencing. The same time-alignment technique was critical in the video quality measurement standard (J.247 PEVQ). PEVQ has been used worldwide as the standard manner that Internet-based video systems have been measured since 08/2008 to improve video delivery services, and is still in use today. A new non-intrusive speech measurement system, developed by Essex in collaboration with Psytechnics, led directly to the standard P.563. This work formed a key product for Netscout and contributed USD500 Million annual revenue to the company by 2018. Through this product, Essex research was used to monitor networks of service providers, government agencies, large financial institutions and other enterprises across the globe to improve the quality of speech communication.

2. Underpinning research

Research from the University of Essex has created new measurements of the quality of speech and video transmitted over the Internet. This work was needed because of the transition of mass media transmission from traditional broadcast TV and telephone systems to the new packet networks (e.g., the Internet), which started around the year 2000, but which is now ubiquitous. This change required completely new methods to measure the quality of speech and video transmission that were pioneered by University of Essex research. In particular, two such developments are described below, the first considered the time alignment of speech and video so that they can be measured accurately in packet-based (Internet) transmission that is inherently asynchronous [R1, R2, R3]. The second innovation considers how a speech transmission system (such as the Internet) can be measured without disturbing the conversation in progress; this is termed a “non-intrusive” measurement [R4]. All of this research was carried out collaboratively with Psytechnics [S2, S7] and appears in peer-reviewed international conferences or journals with more than 150 citations.

Time alignment methods for speech and video quality measuring systems:

A strategic piece of work by the University of Essex, carried out by Reed [R1] and Ghanbari [R2, R3], in collaboration with Psytechnics [S2], investigated the problem of time-alignment in measuring media quality transmitted over the Internet. In Internet streaming, data is split into individual packets that may be subject to arbitrary delay. This was a problem for existing measurement systems that relied upon the received signal being exactly aligned with the original. University of Essex researchers addressed the issue by pioneering practical statistical methods to align the media allowing the measurement of speech and video quality in the new packet-based transmission medium of the Internet. The technique relies upon creating a histogram of audio or video temporal events and using this to compare over previous/later frames to find the optimal alignment point. This work was used by Psytechnics as an essential component of their measurement products [S2] and for the Psytechnics contribution to the ITU-T standard for video quality measurement J.247 [S4] (also called PEVQ); this continues to be a standard method for measuring the quality of encoded video transmitted over packet media such as the Internet.

Non-intrusive speech quality measurement:

The non-intrusive measurement of speech quality carried out by Rob Massara [R4], again in collaboration with Psytechnics [S2, S7], led directly to the ITU-T standard P.563 [S6]. The paper [R4] describes an overall method for performing this new measurement which is described in Figure 8 of the standard [S6], for example Equation 3 in the paper, which models the human vocal tract, is used directly in the standard [S6, Section 9.2.2]. This work is important, as it was the basis of the first practical implementation of a method to measure speech quality without having access to the original signal, i.e., a no-reference measurement. Previous techniques either, required adding disturbing signals during the communication or measuring the communication system (e.g., mobile phone and intervening systems) before or after the actual conversation; this meant it was not possible to measure the quality of any particular conversation while the conversation was in progress. This novel non-intrusive measurement technique models the human vocal tract and compares the measured speech with this model to discriminate between real speech and errors in the transmission. Consequently, the speech quality can be determined by simply monitoring the transmitted voice data, while the conversation is in progress, without disturbing the conversation.

3. References to the research

[can be supplied by HEI on request]

[R1] L. Malfait, P. Gray and M. J. Reed, “Objective listening quality assessment of speech communication systems introducing continuously varying delay (time-warping): A time alignment issue,” 2008 IEEE International Conference on Acoustics, Speech and Signal Processing, Las Vegas, NV, 2008, pp. 4213-4216. doi: 10.1109/ICASSP.2008.4518584

[R2] Q. Huynh-Thu and M. Ghanbari, "No-reference temporal quality metric for video impaired by frame freezing artefacts," 2009 16th IEEE International Conference on Image Processing (ICIP), Cairo, 2009, pp. 2221-2224. doi: 10.1109/ICIP.2009.5413894

[R3] Q. Huynh-Thu and M. Ghanbari, "Temporal Aspect of Perceived Quality in Mobile Video Broadcasting," in IEEE Transactions on Broadcasting, vol. 54, no. 3, pp. 641-651, Sept. 2008.doi: 10.1109/TBC.2008.2001246

[R4] P. Gray, M. P. Hollier and R. E. Massara, "Non-intrusive speech-quality assessment using vocal-tract models," in IEE Proceedings - Vision, Image and Signal Processing, vol. 147, no. 6, Dec. 2000, pp. 493-501. doi: 10.1049/ip-vis:20000539

4. Details of the impact

The work at the University of Essex on the measurement of speech and video quality has had global impact on speech and video streaming services; these types of services contributed to 75% of the Internet traffic by 2017 and growing (estimated as 82% by 2022) [S1]. The first specific impact is the work by Reed [R1] which led to reliability measurement systems incorporated from 2006 until 2016 into the global product Microsoft Skype for Business [S2] with more than 100 Million users [S3]. The work by Ghanbari [R2, R3] built on the work of Reed [R1] to provide time-alignment for video and was included in ITU-T standard J.247, also called PEVQ [S4]; J.247 has been the standard manner that Internet based video systems have been measured since 08/2008 [S5, S4]. The work by Massara [R4] let to the ITU-Standard P.563 [S6], which formed an important product for Psytechnics and later Netscout [S7] where it has been used by global industry to improve the quality of speech communication [S7].

The time alignment work by Reed from University of Essex [R1], in collaboration with Psytechnics [S2], has become an essential part of the systems used by industry to measure and improve Internet speech communication quality and reliability. The findings of this research were incorporated into Microsoft’s Skype For Business product (also called Lync) from 2006 to 2016 [S2]. In 2015 Microsoft stated that there were 100 Million people using Microsoft Skype for Business for their work [S3], and that as ‘ part of the Office 365 platform, Skype for Business will deliver at massive worldwide scale, with datacenters in 37 countries; with the most comprehensive productivity experience; and enterprise-grade reliability and controls on top of a secure and compliance-ready platform’ [S3]. The implementors of the measurement system, formally from Psytechnics, state that the ‘research from the University of Essex in the area of time alignment for the measurement of quality of service of audio and video media was an essential component in the systems developed by Psytechnics. *Specifically, their time alignment method was used in NEXTGENPESQ, PsyVolP products. NEXTGENPESQ and PsyVolP were sold under license in 2006 to Microsoft, and PsyVolP was thus integrated into Lync (later also called Skype for Business) by Microsoft for live monitoring of this product. This was an essential component for monitoring and improving the quality of Microsoft's Lync (Skype for Business) teleconferencing product and the time alignment method originating from the University of Essex's research was used in that product until 2016. Without such a component, Microsoft would not have been able to determine the speech quality of their product. The extent to which Essex's research on improving the quality of teleconferencing was used proved to be truly global as Lync was included in all Enterprise editions of Microsoft Office’ [S2]**. In 2016, an independent report found that 33% of US enterprises (companies with more than 500 employees) were using Lync/Skype for Business [S8, p9].

The time alignment work at the University of Essex by Reed [R1] was further developed by Ghanbari [R2, R3], in collaboration with Psytechnics [S2], for use with video quality measurement and included in the J.247 standard [S2, S4]; Psytechnics’ implementation is described in Annex C of the standard ( [S4], Psytechnics full reference method, p.51 – 79). This objective measurement system, using the time alignment process produced by the University of Essex, replaces time consuming subjective measurements (using large panels of users), thus significantly reducing costs to industry. Opticom, who are the provider of the license of the J.247 measurement system, list 29 large multi-national companies (including Microsoft and Ericsson) from 10 different countries (including US, China, South Korea, Germany) licensed to use PEVQ (i.e. J.247) to test video quality for their products and thus improve their video delivery services [S5].

The research work by the University of Essex into non-intrusive speech quality by Massara [R4], ‘ was an essential component of Psytechnics Experience Manager and Psytechnics no-reference model was integrated into the ITU-T standard P.563’ [S6] [S2]. This product line was sold to Netscout in 2011 [S7] and was subsequently formed (after rebranding) into the Netscout product nGeniusOne in 2015 [S7], which measures the quality of Internet/packet-based speech communication [S9]. In the form of the Netscout nGeniusOne product, the ‘ non-intrusive speech monitoring work, from the Essex collaboration, became an essential differentiator in Netscout nGeniusONE and was one of the reasons customers chose to invest in the solution which had an annual revenue of approximately USD 350 Million by the end of 2016. Netscout's nGeniusONE established itself as an important product for Netscout and by 2018 was responsible for USD 500 Million of annual revenue for Netscout, with the work originating from the research undertaken in collaboration with the University of Essex providing an essential component for the speech monitoring in Netscout nGeniusONE’ [S7] . This innovation from the University of Essex ‘ *continues to be used by mobile and fixed service providers, government agencies, large financial institutions, healthcare providers and a range of other enterprise customers from varied industries across the globe’* [S7] .

5. Sources to corroborate the impact

[S1] Cisco, “Cisco Visual Networking Index Complete Forecast Highlights,” 2018,

https://www.cisco.com/c/dam/m/en_us/solutions/service-provider/vni-forecast-highlights/pdf/Global_2022_Forecast_Highlights.pdf (Accessed 29/1/2021)

[S2] Corroboration from VP/CTO Communications Business Group, Dolby (formally CEO/CTO Psytechnics) that research from the University of Essex was included in the standards J.247 and P.563 and confirmation of the commercial impact of the time-alignment technique in the Microsoft audio conferencing tools Lync/Skype for Business.

[S3] Post from the Skype for Business Team March 18th 2015 https://www.microsoft.com/en-us/microsoft-365/blog/2015/03/18/skype-for-business-is-here-and-this-is-only-the-beginning/ (Accessed 15/1/2021)

[S4] ITU-T Standard J.247 “Objective perceptual multimedia video quality measurement in the presence of a full reference,” August 2008, https://www.itu.int/rec/T-REC-J.247/en

(Accessed 15/1/2021)

[S5] List of PEVQ licensees (list is continuously updated) http://www.pevq.com/licensees.html (Accessed 16/1/2021)

[S6] ITU-T Standard P.563 “Single-ended method for objective speech quality assessment in narrow-band telephony applications,” May 2004 https://www.itu.int/rec/T-REC-P.563/en (Accessed 15/1/2021)

[S7] Corroboration from Assistant Vice President, Product Management, Netscout Systems. Confirming that the work from the University of Essex has been incorporated into NetScout nGeniusOne leading to global use and a substantial revenue stream for NetScout.

[S8] InfoTrack for Unified Communications, “Impact of Microsoft Skype for Business on the Enterprise Voice Market—2016,” September 2016,

https://news.microsoft.com/uploads/2017/01/IUC-MS-Skype4B-2016-Full-Report-10-21-16.pdf (Accessed 15/1/2021)

[S9] Solution Brief from Netscout “Managing Voice, Video Call Quality Issues in Contact Centers with the nGeniusONE Service Assurance Platform,” 2019, https://www.netscout.com/sites/default/files/2019-09/ESB_023_EN-1901%20-%20Managing%20Voice%20Video%20Call%20Quality%20Issues.pdf

(Accessed 11/09/2020)

Submitting institution
The University of Essex
Unit of assessment
11 - Computer Science and Informatics
Summary impact type
Technological
Is this case study continued from a case study submitted in 2014?
No

1. Summary of the impact

Essex research has revolutionised on-chip debug, monitoring and analytics to deliver world leading on-chip diagnostic solutions via commercialisation by UltraSoC Technologies, a company whose creation and growth has been underpinned by Essex research. These solutions ensure correct functionality of complex microchips, substantially accelerate product development and ensure correct operation of the deployed systems in many, often safety critical, domains. Since August 2013, UltraSoC has attracted investment worth over USD 20 million and has licensed this technology to more than 20 leading edge start-ups and Tier 1 semiconductor companies including ARM, Huawei, Microsemi, Intel, Seagate and Western Digital. Employing this technology enables SoC design teams to double their profitability and reduce their design costs by 25%. UltraSoC’s technology is also emerging as the de-facto standard debug support interface for RISC-V processor platform. UltraSoC so successfully enabled semiconductor industry customers to overcome manufacturing defects, software and hardware bugs, device early-failure and wear-out, as well as improve devices’ functional safety and malicious attack protection, that it was acquired by Siemens in 2020 to help the global technology powerhouse provide comparable support.

2. Underpinning research

Since the early 2000s Multiprocessor Systems-on-a-Chip (MPSoC) have become extensively used in electronic systems. Such platforms are now commonplace in everyday life and underpin a vast array of consumer items including cars, smartphones and household appliances. The successful development of these products relies on, on-chip de-bugging and analysis in a short timeframe. However, the advancement of SoC technology, and particularly the move towards MPSoCs, rendered previous software debugging strategies obsolete, unreliable or insufficient. Traditionally these strategies were focused on providing debug support for chips comprising single processors, or multiple processors of a specific family of processor architectures.

Professor Klaus McDonald-Maier identified the need for support of software application development in SoC architectures, especially in cases where complex software is required to interact and execute on multiple processor cores. In cases where SoCs feature other highly interactive blocks (which may contribute to undesired behaviour of the system), this presented a significant technical challenge [R1, R2]. Beginning at the University of Essex in 2005, McDonald-Maier and Hopkins worked on EPSRC funded projects [G1 – G4] that developed initial concepts to address this into practical implementations [R3]. McDonald-Maier’s subsequent work [e.g. G5 and G6] expanded this technology [R4-6].

The research [R3] provided a highly modular debug support architecture, consisting of two important stages. Firstly, debug support adapters were provided in order to connect each processor core, peripheral or interconnect, to the debug infrastructure. The second stage then controlled these adapters, combining their debug data streams in order to preserve timing and compress resulting data to an absolute minimum. Critically, this compression meant that debug data could be straightforwardly sent from the SoC to an external development station or PC using a variety of limited bandwidth interfaces. This is an important characteristic because every product chip includes this additional debug infrastructure and therefore it is required in a high volume of devices.

The development of this process represented the first systems-centric debug support architecture for SoCs featuring multiple processor cores (i.e. cores from multiple IP provider such as ARM, MIPS etc. as well as other active peripherals). The architecture substantially outperformed the state of the art (i.e. by an order of magnitude) and, notably, achieved this in a significantly more compact implementation than existing architectures. The developed process provides debug support for two processor cores using less logic than that required for one processor core when using previous state of the art. Core contributions focus on ease of integration of processors and other components from different vendors and improved detection capability for unusual events and capability to provide this during the deployment stage of the respective SoC and for cybersecurity purposes.

Further work from 2014 focused on increasing ease of integration and flexibility with a message-based infrastructure and focus on detection capability for systems security and analytics [R4-5]. The Essex research group continued to contribute to the development of this technology in UltraSoC, the University of Essex spinout established for its commercialisation on it, and Siemens, who acquired it. For instance, as part of EPSRC funded research projects SPIRIT [G5] and the EPSRC National Centre for Nuclear Robotics [G6], there was extensive direct collaboration, focusing on methods for on-line detection of faults introduced through exposure to extreme environments such as radiation [R6].

3. References to the research

[can be supplied by HEI on request]

[R1] Hopkins, A.B.T. and K.D. McDonald-Maier (2006) Debug support for complex Systems on-Chip: A review, IEE Proceedings on Computers and Digital Techniques, 153(4), 197-207. (212 citations – July 2019) DOI:10.1049/ip-cdt:20050194

[R2] Mayer, A., H. Siebert and K.D. McDonald-Maier (2007) Boosting Debug Support for Complex Systems-on-Chip, IEEE Computer, 40(4), 76-81. (44 citations – July 2019) DOI:10.1109/MC.2007.118

[R3] Hopkins, A.B.T. and K.D. McDonald-Maier (2006) Debug Support Strategy for Systems-on-Chips with Multiple Processor Cores, IEEE Transactions on Computers, 55(2), 174-184. (100 citations – July 2019) DOI:10.1109/TC.2006.22

[R4] Zhai, X., K. Appiah, S. Ehsan, G. Howells, H. Hu, D. Gu, K. McDonald-Maier (2015) A Method for Detecting Abnormal Program Behaviour on Embedded Devices, IEEE Transactions on Information Forensics and Security, 10(8), 1692-1704, DOI 10.1109/tifs.2015.2422674

[R5] Alheeti KMA, Al-ani, MS, McDonald-Maier K (2018) A hierarchical detection method in external communication for self-driving vehicles based on TDMA. PLoS ONE 13(1): e0188760. https://doi.org/10.1371/journal.pone.0188760

[R6] Saha, S., Ehsan, S., Stoica, A., Stolkin, R. and K. McDonald-Maier (2018) Real-Time Application Processing for FPGA-Based Resilient Embedded Systems in Harsh Environments, 2018 NASA/ESA Conference on Adaptive Hardware and Systems (AHS), Edinburgh, 2018, 299-304. doi: 10.1109/AHS.2018.8541449

Research funding:

[G1] McDonald-Maier, Debug support strategy for systems-on-chips with multiple processor cores, (EPSRC), Aug ‘05 – May ‘06, £39,218

[G2] McDonald-Maier, ReSIP – Reconfigurable system-on-chip based networks of integrated and distributed sensor platform nodes for environmental diagnostic and sensing, (EPSRC), Oct ‘05 – Sep ‘08, £265,844

[G3] McDonald-Maier, Networking of distributed sensors for proactive condition monitoring of wind, (EPSRC), Oct ‘05 – Jan ‘09, £213,374

[G4] McDonald-Maier, ESPACENET – Evolvable networks of intelligent and secure integrated and distributed reconfigurable system-on-chip sensor nodes for aerospace based monitoring and diagnostics, (EPSRC), Oct ‘05 – Nov ‘08, £268,856

[G5] McDonald-Maier, SPIRIT, (EPSRC), Jan ’17 – Dec ’19, £236,494;

[G6] McDonald-Maier, National Centre for Nuclear Robotics (NCNR), (EPSRC), Oct ’17 – Mar ’21, £11,588,431, of which £1,386,737.010 was for Essex

4. Details of the impact

UltraSoC was spun out to commercialise the debug and on-chip analytics technology invented as part of EPSRC funded research [G1-G4]. The company has subsequently been built around research undertaken by McDonald-Maier and his Embedded and Intelligent Systems (EIS) laboratory at the University of Essex [S1]. The research group at Essex recognised that the outputs of their work [R1-R3] held broad applicability for providing debug support to MPSoCs in a vast array of global scenarios that rely on embedded systems. In a marketplace where nearly half the cost of chip development was spent on de-bugging activities, the novel architecture resulting from the research conducted at Essex enables development of more reliable software, having significant economic and safety implications in consumer electronics and safety-critical applications [S1] (self-driving vehicle electronics exemplify this). Essex researchers sought to share the capabilities and associated benefits of this technology with a wide audience and developed a robust strategy in order to transform research insight into practical benefit. This centred on a broad range of dissemination activities that targeted investment from a variety of sources, in order to help commercialise the technology via its spin-out UltraSoC Technologies.

UltraSoC has developed significantly since August 2013 with Venture capital funding raised from the original series A investors Octopus and new investors led by electronic design legend and Chairman of UltraSoC, Prof Alberto Sangiovanni-Vincentelli (UBerkeley and co-founder of Cadence and Synopsis, Atlante Ventures) and expansion to new design centres in Bristol and Poland, increasing the number of employees from 12 in July 2013 to over 40 in 2020 (FTEs: 40) [S1]. UltraSoC raised significant rounds, with GBP 5M led by Atlante for the continued expansion of the UltraSoC team and its product portfolio to support all mainstream embedded processor platforms in 2017 (see C21 of [S2]) and UltraSoC secured new investment of GBP 5 M (see C21 of [S2]) and in 2019 an additional GBP 5M to focus on hardware security (see C9 of [S2]) . Recently, UltraSoC has revisited the work on analytics, originally undertaken in the RESIP [G2] and Espacenet [G4] grants, where this was employed for design space optimisation and security applications and expanded this towards applications in AI and Machine Learning [R4]. In June 2020, UltraSoC was acquired by Siemens [S6] to enable semiconductor industry customers overcome key pain points including manufacturing defects, software and hardware bugs, device early-failure and wear-out, functional safety, and malicious attacks [S1] and [S3]. [S1] confirms:

‘The research undertaken by you and your EIS laboratory at the University Essex has provided the

technological foundations of the UltraSoC product portfolio and enabled us to build practical debug and analytic solutions which proved to be of such immense commercial and technological value they established UltraSoC as the world leading provider in the debug analytics and cybersecurity technology market. Consequently, UltraSoC was the company Siemens chose to acquire to achieve a step change in its activities in this market.’ [S1]

UltraSoC’s embedded analytics IP and debugging tools are used to monitor and boost the performance, reliability and safety of consumer electronics, safety-critical vehicle electronics, AI chips, servers and high performance computing platforms. Early customer PMC-Sierra, the fabless semiconductor company (acquired by Microsemi in 2016), used analytics and other monitoring tools within its disk drive controllers, which power a significant proportion of server chipsets globally. The monitors were used to collect detailed data on chip behaviour, while also shedding light on the performance of server infrastructure those SoCs support. “The hardware-based approach can detect hard-to-identify issues [making] it substantially easier to home in on non-fatal bugs.” (see C23 of [S2]).

Extensive licensing activity followed this breakthrough with PMC-Sierra, with UltraSoC’s technology increasingly seen as the world leading on-chip analytics capability to enable companies to bring their products to market rapidly, resulting in a series of technology licenses to global companies like ARM (2016 see C20 of [S2]), HiSilicon (Huawei, 2016), Esperanto (2017, see C18 of [S2]), Mircosemi (2017), Movidius (Intel, 2016), Alibaba (2018, see C17 of [S2]), Kraftway (2018, see C15 of [S2]), Seagate (2019, see C9 of [S2]) and Western Digital (2019 see C9 of [S2]). UltraSoC achieved 100% client retention; all licensees continued their engagement with UltraSoC, renewing and expanding their licenses. This was facilitated by its universal monitoring and analytic system to support rapid ‘plug and play’ integration of semiconductor IP blocks from different vendors (e.g. processors from ARM, MIPS, RISC-V etc.), enabling to effectively support mixed IP SoCs with a minimum of engineering effort (see C29 of [S2]) and fully support this via respective tools integration through its partners Andes, Arm, Cadence/Tensilica, Imperas, Lauterbach, Mentor, Percepio, Segger, SiFive, and Sondrel [S1].

UltraSoC’s technology has had a distinct impact on the economics of the semiconductor industry, where its “intelligent analytics’ closed the productivity gap created by the failure of traditional SoC development methodologies to keep pace with escalating systemic complexity. Providing engineering teams actionable insights that shorten the total development cycle time, accelerate debug, and reduce risk and cost to ensure timely market entry. Analysis from SemiCo research demonstrates the bottom-line value of this approach – SoC design teams can double their profitability and reduce their development costs by a quarter by using UltraSoC [S1]. Thus, UltraSoC’s in chip monitoring is increasingly seen as an essential part of coping with “rising complexity and a spectrum of possible interactions” (see C23 of [S2]) which is particularly vital for the monitoring and analytics for automotive and safety-critical systems, increasingly important for self-driving and autonomous systems (see C13 of [S2]), where it is necessary to “adopt a new approach of looking inside the electronics.” With “On-chip monitoring … allowing continuous measurement of previously-inaccessible information … so that users can actually take corrective action at every stage.” (see C25 of [S2]). These on-chip monitoring mechanism also provide key capabilities for systems level cybersecurity through its Lockstep Monitor (See C22 of [S2]) and its Sentinel technology [S5 and S6], this is also evidenced by UltraSoCs selection into the major DARPA Automatic Implementation of Secure Silicon programme to enable “scalable defence mechanisms into chip designs” (see C30 of [S2]).

Additionally, UltraSoC provides the only commercial grade debug solution for emerging ARM alternative Open Source processor core platform RISC-V (see C1 and C2 of [S2]) with its “standards-compliant RISC-V trace solution is a major contribution … to create a comprehensive ecosystem that delivers robust, commercial grade open-source platforms” (C1 of [S2]).

UltraSoC’s development of novel chip design technology which addresses the big problem of complex Integrated circuit design was recognised at the 2015 Elektra Awards. At the European ‘Oscars’ of the electronic Industry, UltraSoC was named Best New Company. The judges were impressed by its partnerships with leading firms for its patented IC debug tools, which are already making an impact in global semiconductor markets. [S4] Its cybersecurity focused version of the technology was recognised with selection as finalist for the Embedded Solution Product of the Year in the 2020 Electronic Industry Awards [S5] and with the Best in Show Security Award at the globally leading Embedded World 2020 trade show [S6].

Finally, the General Manager, Siemens Digital Industries Software and Tessent Vice President confirms: “ Siemens’ acquisition of UltraSoC means that for the first time our customers can access not just design-for-test, but a comprehensive ‘Design for Lifecycle Management’ solution for system-on-chips, including functional safety, security and optimization,” Adding: “ By utilizing design augmentation to detect, mitigate and eliminate risks throughout the SoC lifecycle, customers can radically improve time-to-revenue, product quality & safety, and profitability. UltraSoC has a fast-growing business and impressive customer list and, as part of Siemens, can complement Tessent to create a truly unique offering in the market.” [S3]. Siemens also confirms that “ UltraSoC is a pioneer of embedding monitoring hardware into complex SoCs to enable “fab-to-field” analytics capabilities designed to accelerate silicon bring-up, optimize product performance, and confirm that devices are operating “as designed” for functional safety and cybersecurity purposes.” [S3]. All this from the company with a product portfolio founded on, and practical debug and analytic solutions enabled by research led by McDonald-Maier at the University Essex [S1].

5. Sources to corroborate the impact

[S1] Former Chief Strategy Officer, UltraSoC Technologies now Senior Director Portfolio Strategy at Mentor, a Siemens Business

[S2] Compilation of links from open access press and media publications evidencing company developments and significance

[S3] Compilation of links covering UltraSoC acquisition by Siemens (e.g. ‘Siemens acquires UltraSoC to drive design for silicon lifecycle management’)

[S4] 2015 Elektra Awards: UltraSoC named Best New Company: https://www.electronicsweekly.com/news/elektra-awards-2015-the-winners-2015-11/

[S5] Finalist for the Embedded Solution Product of the Year in the 2020 Electronic Industry Awards https://electronicsindustryawards.co.uk/finalists/

[S6] UltraSoC wins Security Award https://www.realwire.com/releases/UltraSoC-wins-Security-Award-for-Bus-Sentinel-hardware-cybersecurity-IP

Showing impact case studies 1 to 6 of 6

Filter by higher education institution

UK regions
Select one or more of the following higher education institutions and then click Apply selected filters when you have finished.
No higher education institutions found.
Institutions

Filter by unit of assessment

Main panels
Select one or more of the following units of assessment and then click Apply selected filters when you have finished.
No unit of assessments found.
Units of assessment

Filter by continued case study

Select one or more of the following states and then click Apply selected filters when you have finished.

Filter by summary impact type

Select one or more of the following summary impact types and then click Apply selected filters when you have finished.

Filter by impact UK location

UK Countries
Select one or more of the following UK locations and then click Apply selected filters when you have finished.
No UK locations found.
Impact UK locations

Filter by impact global location

Continents
Select one or more of the following global locations and then click Apply selected filters when you have finished.
No global locations found.
Impact global locations

Filter by underpinning research subject

Subject areas
Select one or more of the following underpinning research subjects and then click Apply selected filters when you have finished.
No subjects found.
Underpinning research subjects