HUMAN-AGENT COLLECTIVES: FROM FOUNDATIONS TO APPLICATIONS [ORCHID]
Lead Research Organisation:
University of Southampton
Department Name: Electronics and Computer Science
Abstract
With a reported 5 billion mobile subscriptions worldwide, access to communication technologies has reached unprecedented levels and has fundamentally altered the ways in which we experience computational systems. Once delivered through a desktop machine to an office worker, computing has become an interwoven feature of everyday life across the globe in a way that profoundly affects us all. We are now interconnected using mobile devices; we routinely invoke remote services through a global cloud infrastructure and increasingly rely on computational devices in our everyday life. Computational devices monitor our health, entertain us, guide us and keep us safe and secure. However, this explosive growth in these devices and on-line services is only a precursor to an era of ubiquity, where each of us will routinely rely upon a plethora of smart and proactive computers that we carry with us, access at home and at work, and that are embedded into the world around us. As computation increasingly pervades the world around us, it will profoundly change the ways in which we work with computers. Rather than issuing instructions to passive machines, we will increasingly work in partnership with highly inter-connected computational components (aka agents) that are able to act autonomously and intelligently. Specifically, humans and software agents will continually and flexibly establish a range of collaborative relationships with one another, forming human-agent collectives (HACs) to meet their individual and collective goals. This vision of people and computational agents operating at a global scale offers tremendous potential and, if realised correctly, will help us meet the key societal challenges of sustainability, inclusion, and safety that are core to our future. However, these benefits are mirrored by the potential of equally concerning pitfalls as we shift to becoming increasingly dependent on systems that interweave human and computational endeavour.As systems based on human-agent collectives grow in scale, complexity and temporal extent, we will increasingly require a principled science that allows us to reason about the computational and human aspects of these systems if we are to avoid developments that are unsafe, unreliable and lack the appropriate safeguards to ensure societal acceptance.Delivering this science is the core research objective of this Programme. In more detail, it seeks to establish the new science that is needed to understand, build and apply HACs that symbiotically interleave human and computer systems to an unprecedented degree. To this end, it brings together three world-leading academic groups from the Universities of Southampton, Oxford and Nottingham (with multi-disciplinary expertise in the areas of artificial intelligence, agent-based computing, machine learning, decentralised information systems, participatory systems, and ubiquitous computing) with industrial collaborators (initially BAE Systems, PRI Ltd and the Australian Centre for Field Robotics) to collectively establish the foundational scientific underpinnings of these systems and drive these understandings to real-world applications in the critical domains of future energy networks, and disaster response.
Planned Impact
A key component of ORCHID is the close connection and inter-play between world-leading fundamental research, the demonstration of this research in compelling real-world application scenarios, and the involvement of collaborating partners whose future lies in exploiting such research in these application areas. This provides the ideal environment for producing high impact research across a broad set of areas including: - Advances in knowledge of direct relevance to those involved in understanding the design, construction and use of systems that exploit HACs. These advances will be reflected in new concepts and theories, new interactive techniques and new platforms to support human-agent collaboration (i.e. the new science of HACs). - Novel technologies and demonstrators that showcase the use of HACs in a number of application areas which will show which methods, models and technologies are effective. In particular, the proof of concepts, working prototypes, simulations and studies will provide empirical evidence of the theoretical concepts, the technical feasibility of their construction and the social acceptability of the technologies. - Open source sets of code, toolkits and competitions that encourage a wider community to share and contribute to the research. This will target a broad set of industrial research labs and advanced developers exploring the ways in which increasingly large scale agent arrangements might be exploited. Particularly, relevant in this regard is the Horizon Digital Economy hub with which we will co-promote events and so gain access to their extensive network of partners. This work is of potential interest to a broad set of beneficiaries. In particular, the focus on reducing environmental impact seeks to enhance the quality of life for the everyday citizen and the planet and the work on disaster response seeks to exploit technology to rescue and preserve life. In order to identify where research can be applied, and also for identifying new areas for HAC systems and new collaboration partners, a dedicated Knowledge Transfer Officer will be employed. As part of ORCHID's regular management process, we will reflect upon and update the knowledge transfer strategy by reviewing the key target audiences and the mechanisms used to deliver to them. Our initial set of communities to engage include: - Our collaborating partners, who will benefit from access to the leading researchers in this new area, the ability to shape the ongoing research as it happens, the exchange of researchers through inward and outward secondments, and the ability to demonstrate and evaluate the work in scenarios that closely align with their applications of interest. In recognition of this fact, all of these partners will make a significant investment, which at present, totals over 2M. - Charities concerned with promoting the reduction of carbon emissions and response to disasters, such as the Carbon Trust, Energy Saving Trust, Ushahidi, and the Disasters Emergency Committee, who will benefit from the abilities of HAC technologies to aid in energy reduction and disaster response and the ability to more directly involve people in helping with these issues. - Policy makers and the general public who will benefit from an understanding of the capabilities and issues to emerge from human-agent collectives. For example, the shift to increasingly open access to data, through initiatives such as data.gov.uk, has highlighted innovation in this space and placed issues of access, availability and control of data on the political agenda. The next decade will require a still closer partnership between researchers and law-makers as we establish the key principles of governance in this domain, and we will exploit our links with the James Martin 21st Century School to do so.
Publications
Rahwan T
(2015)
Coalition structure generation: A survey
in Artificial Intelligence
Bachour K
(2015)
Provenance for the People
Wu F
(2015)
Agile Planning for Real-World Disaster Response
Ramchurn SD
(2015)
HAC-ER: A disaster response system based on human-agent collectives.
Karastergiou A
(2015)
Limits on fast radio bursts at 145 MHz with artemis, a real-time software backend
in Monthly Notices of the Royal Astronomical Society
Rajpaul V
(2015)
Ghost in the time series: no planet for Alpha Cen B
in Monthly Notices of the Royal Astronomical Society: Letters
C. Lloyd
(2015)
Inferring Dynamic Interaction Networks with N-LPPA
Pike M
(2015)
Two-way affect loops in multimedia experiences
Tran-Thanh L
(2015)
Crowdsourcing Complex Workflows under Budget Constraints
Lundgren S
(2015)
Designing Mobile Experiences for Collocated Interaction
Rajpaul V
(2015)
A Gaussian process framework for modelling stellar activity signals in radial velocity data
in Monthly Notices of the Royal Astronomical Society
Panagopoulos AA
(2015)
Towards optimal solar tracking: a dynamic programming approach
Psorakis I
(2015)
Inferring social structure from temporal data
in Behavioral Ecology and Sociobiology
Moreau L
(2015)
The rationale of PROV
in Journal of Web Semantics
Sezavar Keshavarz A
(2015)
Provenance and Annotation of Data and Processes
Ramchurn S
(2015)
Human-agent collaboration for disaster response
in Autonomous Agents and Multi-Agent Systems
Moreau L.
(2015)
Provenance of publications: A PROV style for latex
in 7th USENIX Workshop on the Theory and Practice of Provenance, TaPP 2015
Michalak T
(2014)
Implementation and Computation of a Value for Generalized Characteristic Function Games
in ACM Transactions on Economics and Computation
Wu F
(2014)
Regret-based multi-agent coordination with uncertain task rewards
in Proc. 28th Conf. on AI (AAAI)
Moran S
(2014)
Literatin
Han T
(2014)
The cost of interference in evolving systems
in Proceedings of COIN 2014: The 17th International Workshop on Coordination, Organisations, Institutions and Norms
Naroditskiy V
(2014)
Crowdsourcing contest dilemma.
in Journal of the Royal Society, Interface
Costanza E
(2014)
Doing the laundry with agents
Long Tran-Thanh (Author)
(2014)
BudgetFix: budget limited crowdsourcing for interdependent task allocation with quality guarantees
Fischer J
(2014)
Energy advisors at work
Venanzi M
(2014)
Community-based bayesian aggregation models for crowdsourcing
Tran-Thanh L
(2014)
Efficient regret bounds for online bid optimisation in budget-limited sponsored search auctions
in Proceedings of 30th Conference on Uncertainty in Artificial Intelligence
Moreau L
(2014)
Fundamental Approaches to Software Engineering
De Clippel G
(2014)
Destroy to save
in Games and Economic Behavior
Naroditskiy V
(2014)
Referral incentives in crowdfunding
in Proceedings of HCOMP2014: Conference on Human Computation & Crowdsourcing
D. T. Huynh
(2014)
Data quality assessment from provenance graphs
A. Levenburg
(2014)
Predicting Economic Indicators from Web Text Using Sentiment Composition
Biermann F
(2014)
Task assignment with controlled and autonomous agents
in Mathematical Social Sciences
S. Reece
(2014)
Efficient State-Space Inference of Periodic Latent Force Models
in Journal of Machine Learning Research
J. E. Fischer
(2014)
Supporting Team Coordination on the Ground: Requirements from a Mixed Reality Game
Alan A
(2014)
A field study of human-agent interaction for electricity tariff switching
in Proceedings of the 13th International Conference on Autonomous Agents and Multi-Agent Systems
Tran-Thanh L
(2014)
BudgetFix: budget limited crowdsourcing for interdependent task allocation with quality guarantees
in AAMAS '14 Proceedings of the 2014 international conference on Autonomous agents and multi-agent systems
S. Reece (Author)
(2014)
Efficient state-space inference of periodic latent force models
in Journal of Machine Learning Research
Parson O
(2014)
An unsupervised training method for non-intrusive appliance load monitoring
in Artificial Intelligence
Tran-Thanh L
(2014)
Efficient crowdsourcing of unknown experts using bounded multi-armed bandits
in Artificial Intelligence
Jennings N
(2014)
Human-agent collectives
in Communications of the ACM
Title | ORCHID Artist in Residence |
Description | Film based around ORCHID's contribution to disaster response |
Type Of Art | Film/Video/Animation |
Year Produced | 2016 |
Impact | An exhibition is planned by the artists (Steve Beard and Victoria Halford) in the near future. |
Description | The ORCHID team all shared the ethos of undertaking fundamental research in real-world contexts and are worked closely with the Programme's industrial collaborators to demonstrate and evaluate this fundamental work in application domains of energy systems, citizen science and disaster response. In total 19 Human Agent Collective (HAC) applications were built, including: HAC-ER, a prototype disaster response system that comprises a multi-UAV coordination system, a crowdsourcing component to gather information from inhabitants of an affected area and create heatmaps of the disaster zone, the AtomicORCHID coordination component, and a provenance tracking tool AgentSwitch, a personalised energy tariff recommender system, which incorporates a predication algorithm for yearly energy consumption and disaggregation of energy feeds and Cicada app, an iPhone and Android app, originally developed to detects the New Forest Cicada and subsequently extended to detect other insects. |
Exploitation Route | ORCHID engaged with Rescue Global, Hampshire County Council and Hampshire Fire and Rescue Services around policy and process in disaster response. Work with Rescue Global to develop an augmented bird-table is ongoing. ORCHID research on classifier combination in the citizen science domain was successfully applied to three industrial application problems in collaboration with BAE Systems. Joulo was deployed and tested in collaboration with major energy companies, and subsequently acquired by Quby in 2015. ORCHID engaged with uSwitch who facilitated access to their energy comparison API in support of the design and demonstration of ORCHID's AgentSwitch demonstration system. Other ORCHID supported achievements include: the CrowdScanner system that was developed for, and won, the US State Department's TAG challenge for social mobilisation and rapid information gathering the Bayesian Classifier Combination algorithm for the Shared Task Challenge, developed in collaboration with a Microsoft Research/Bing team, that won the Shared Task Challenge in the Crowdsourcing at Scale workshop an automated fantasy football manager that has outperformed 2.5m players in the online Fantasy Premier League and the OutrunCancer platform that was launched to evaluate networked incentive schemes in collaboration with Cancer Research UK. |
Sectors | Aerospace Defence and Marine Digital/Communication/Information Technologies (including Software) Electronics Energy Environment |
URL | http://www.orchid.ac.uk/ |
Description | ORCHID has released software both to support academic articles (Latent force software accompanying JMLR article), in collaboration with industry (Infer.NET which is a MSR software that automatically generates large scale inference code for Bayesian graphical models) and as open access toolkits (Prov Python library version 1.1.0 (https://pypi.python.org/pypi/prov) which has over 46,000 downloads, the ProvStore (https://provenance.ecs.soton.ac.uk/store/): a public online repository for provenance and NILMTK: an open source non-intrusive load monitoring tool kit. |
First Year Of Impact | 2012 |
Sector | Aerospace, Defence and Marine,Digital/Communication/Information Technologies (including Software),Energy,Environment |
Impact Types | Societal Economic |
Description | Improving local authority planning |
Geographic Reach | Local/Municipal/Regional |
Policy Influence Type | Citation in other policy documents |
Impact | The HCC emergency planning unit used our software to better plan their response to the mass evacuation of civilians from homes and businesses over a large region (city scale) such as the one around the Fawley refinery. This has, in turn, led to change in operating procedures. In particular, for example, they have significantly improved their techniques to estimate evacuation completion times, from an educated guess of one hour, that we have shown to be highly optimistic, to a more realistic figure of three hours for Fawley village. These more realistic estimates have helped to better identify choke points in the road network and new congestion control strategies to reduce evacuation times have subsequently been devised and adopted. Target Audience: Local Government |
Description | Police Workshop on Provenance |
Geographic Reach | Local/Municipal/Regional |
Policy Influence Type | Influenced training of practitioners or researchers |
Impact | Workshop was provided to NCA, South East Regional Cyber Crime Unit |
Description | A robust toolbox for exoplanet analysis |
Amount | £220,000 (GBP) |
Organisation | The Leverhulme Trust |
Sector | Charity/Non Profit |
Country | United Kingdom |
Start |
Description | ASUR: Human-agent collaboration for multi-UAV Control |
Amount | £50,000 (GBP) |
Organisation | Defence Science & Technology Laboratory (DSTL) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Adaptive Demand Response |
Amount | £250,000 (GBP) |
Organisation | TSB Bank plc |
Sector | Private |
Country | United Kingdom |
Start |
Description | Aperio: Low cost facade management in naturally ventilated buildings |
Amount | £494,000 (GBP) |
Funding ID | EP/L024608/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Automated assessment of broiler chicken welfare |
Amount | £738,000 (GBP) |
Funding ID | BB/K001388/1 |
Organisation | Biotechnology and Biological Sciences Research Council (BBSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Autonomous behaviour and learning in an uncertain world |
Amount | £1,200,000 (GBP) |
Funding ID | EP/J012300/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Biosound/Mosquito Project |
Amount | £500,000 (GBP) |
Organisation | |
Sector | Private |
Country | United States |
Start |
Description | Casma |
Amount | £500,000 (GBP) |
Organisation | Economic and Social Research Council |
Sector | Public |
Country | United Kingdom |
Start |
Description | CharIoT: Leveraging the Internet of Things to reduce fuel poverty |
Amount | £277,000 (GBP) |
Funding ID | EP/L02392X/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Creating energy for change |
Amount | £1,076,000 (GBP) |
Funding ID | EP/K002589/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Cyber security solutions for smart traffic control systems |
Amount | £500,000 (GBP) |
Funding ID | EP/N02026X/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start | 03/2016 |
End | 02/2018 |
Description | Design by privacy |
Amount | £50,000 (GBP) |
Organisation | Microsoft Research |
Sector | Private |
Country | Global |
Start |
Description | Domestic energy management |
Amount | £15,000 (GBP) |
Organisation | Higher Education Innovation Funding (HEIF) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Future everyday interaction with the Autonomous IoT |
Amount | £806,000 (GBP) |
Funding ID | EP/N014243/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Home Hub of all things as platform for multi sided market powered by Internet of Things |
Amount | £982,000 (GBP) |
Funding ID | EP/K039911/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Human mobility analysis and anomaly detection |
Amount | £77,000 (GBP) |
Organisation | Defence Science & Technology Laboratory (DSTL) |
Department | Centre for Defence Enterprise |
Sector | Public |
Country | United Kingdom |
Start |
Description | Intelligient SME energy management and trading with ancillary services |
Amount | £200,000 (GBP) |
Funding ID | EP/M507167/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | KTS with BAE Systems |
Amount | £50,000 (GBP) |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Mixed reality game pattern cards |
Amount | £10,000 (GBP) |
Organisation | University of Nottingham |
Sector | Academic/University |
Country | United Kingdom |
Start |
Description | Multi-agent collectives for sensing, autonomy, intelligence and control |
Amount | £1,316,000 (GBP) |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | ORCHID Artist in Residence |
Amount | £15,000 (GBP) |
Organisation | The Leverhulme Trust |
Sector | Charity/Non Profit |
Country | United Kingdom |
Start |
Description | Prototyping open innovation models for ICT-enabled manufacturing in food and packaging |
Amount | £1,827,000 (GBP) |
Funding ID | EP/K014234/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Referrak incentives in crowdsourcing |
Amount | £8,000 (GBP) |
Organisation | British Council |
Sector | Charity/Non Profit |
Country | United Kingdom |
Start |
Description | Researcher Links Workshop for Disaster Response |
Amount | £50,000 (GBP) |
Organisation | British Council |
Department | British Council Researcher Links |
Sector | Public |
Country | United Kingdom |
Start |
Description | Thermomodelling and inference with Home OS |
Amount | £60,000 (GBP) |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start |
Description | Transforming feedback - interactive, practice-level visualisation of electricity consumption |
Amount | £50,000 (GBP) |
Organisation | Kingston University London |
Sector | Academic/University |
Country | United Kingdom |
Start |
Description | User centred networking |
Amount | € 5,000,000 (EUR) |
Funding ID | 611001 |
Organisation | European Commission |
Sector | Public |
Country | European Union (EU) |
Start |
Description | biosound |
Amount | £225,000 (GBP) |
Organisation | University of Oxford |
Department | Oxford Martin School |
Sector | Academic/University |
Country | United Kingdom |
Start |
Description | ACFR |
Organisation | University of Technology Sydney |
Country | Australia |
Sector | Academic/University |
PI Contribution | Provision of UAVs and algorithms for field trials |
Collaborator Contribution | Provision of UAV field trial facilities |
Impact | Footage of field trials |
Start Year | 2011 |
Description | Rescue Global |
Organisation | Rescue Global |
Country | United Kingdom |
Sector | Charity/Non Profit |
PI Contribution | provision of ORCHID technologies |
Collaborator Contribution | Field testing of technologies |
Impact | Presentations at AIDR 2014 and development of ORCHID disaster response technologies |
Start Year | 2013 |
Description | Secure |
Organisation | Secure Meters |
Country | India |
Sector | Private |
PI Contribution | Development of smart meter technology |
Collaborator Contribution | Provision of equipment and internship opportunities for students |
Impact | Student internships |
Start Year | 2011 |
Title | A method for inferring a relevance for a data source using regression technique (on kernelised data) |
Description | A method for inferring a relevance for a data source using regression technique (on kernelised data) |
IP Reference | GB1109210.3 |
Protection | Patent granted |
Year Protection Granted | 2012 |
Licensed | Commercial In Confidence |
Impact | Utilisation by BAE Systems |
Title | A method for inferring an output distribution using Gaussian Process (on kernelised data) |
Description | A method for inferring an output distribution using Gaussian Process (on kernelised data) |
IP Reference | GB1109209.5 |
Protection | Patent granted |
Year Protection Granted | 2012 |
Licensed | Commercial In Confidence |
Impact | Utilisation by BAE Systems |
Title | A method for inferring the relevance for a pre-processing technique using machine learning |
Description | A method for inferring the relevance for a pre-processing technique using machine learning |
IP Reference | GB1109206.1 |
Protection | Patent granted |
Year Protection Granted | 2012 |
Licensed | Commercial In Confidence |
Impact | Utilisation by BAE Systems |
Title | Adaptive Task Assignment |
Description | Adaptive task assignment for crowdsourcing platform |
IP Reference | MS355469.01 |
Protection | Patent granted |
Year Protection Granted | 2015 |
Licensed | Commercial In Confidence |
Impact | The ActiveCrowdToolkit .NET v0.1 includes the methods: Majority voting, Vote distribution, Dawid&Skene, Bayesian Classifier Combination (BCC), Community-Based BCC. |
Title | Crowdsourcing system with community learning |
Description | Crowdsourcing system with community learning |
IP Reference | MS 340522.01 |
Protection | Patent granted |
Year Protection Granted | 2015 |
Licensed | Commercial In Confidence |
Impact | The ActiveCrowdToolkit .NET v0.1 includes the methods: Majority voting, Vote distribution, Dawid&Skene, Bayesian Classifier Combination (BCC), Community-Based BCC. |
Title | HETEROGENEOUS DATA FUSION USING GAUSSIAN PROCESSES |
Description | A method and apparatus for processing data, the data comprising: a set of one or more system inputs; and a set of one or more system outputs; wherein each system output corresponds to a respective system input; each system input comprises a plurality of data points, such that at least one of these data points is from a different data source (8, 10, 12, 16, 18, 20, 22) to at least one other of those data points, the method comprising: performing a kernel function on a given system input from the data and a further system input to provide kemelised data; and inferring a value for further system output corresponding to the further system input; wherein the step of inferring comprises applying a Gaussian Process to the kemelised data. The data sources (8, 10, 12, 16, 18, 20, 22) may be heterogeneous data sources. |
IP Reference | WO2012164243 |
Protection | Patent granted |
Year Protection Granted | 2012 |
Licensed | Commercial In Confidence |
Impact | Utilisation by BAE Systems |
Title | ActiveCrowd Toolkit |
Description | The ActiveCrowdToolkit .NET v0.1 includes the methods: Majority voting, Vote distribution, Dawid&Skene, Bayesian Classifier Combination (BCC), Community-Based BCC. |
Type Of Technology | New/Improved Technique/Technology |
Year Produced | 2014 |
Impact | Test crowd consensus methods with one line of code Use a set of existing data aggregation models to combine crowd labels and learn information about the workers. |
URL | http://orchidproject.github.io/active-crowd-toolkit/ |
Title | Gridcarbon |
Description | GridCarbon presents a summary of the generation mix data broken down into major categories of fuel type and various interconnectors. This data is converted into a grid carbon intensity value by weighting each generation type by its contribution to total generation and by its individual carbon intensity. The resulting figure is then divided by 0.93 in order to reflect the 7% losses in the transmission and distribution networks. |
Type Of Technology | Software |
Year Produced | 2012 |
Open Source License? | Yes |
Impact | Increased reputation of ORCHID project. |
URL | http://www.gridcarbon.uk/ |
Title | Latent force software |
Description | Latent force software accompanying JMLR article |
Type Of Technology | New/Improved Technique/Technology |
Year Produced | 2014 |
Impact | Latent force models (LFM) are principled approaches to incorporating solutions to differential equations within non- parametric inference methods. Unfortunately, the development and application of LFMs can be inhibited by their computational cost, especially when closed-form solutions for the LFM are unavailable, as is the case in many real world problems where these latent forces exhibit periodic behaviour. Given this, we develop a new sparse representation of LFMs which considerably improves their computational efficiency, as well as broadening their applicability, in a principled way, to domains with periodic or near periodic latent forces. Our approach uses a linear basis model to approximate one generative model for each periodic force. We assume that the latent forces are generated from Gaussian process priors and develop a linear basis model which fully expresses these priors. We apply our approach to model the thermal dynamics of domestic buildings and show that it is effective at predicting day-ahead temperatures within the homes. We also apply our approach within queueing theory in which quasi-periodic arrival rates are modelled as latent forces. In both cases, we demonstrate that our approach can be implemented efficiently using state-space methods which encode the linear dynamic systems via LFMs. Further, we show that state estimates obtained using periodic latent force models can reduce the root mean squared error to 17% of that from non-periodic models and 27% of the nearest rival approach which is the resonator model (Sarkka et al., 2012; Hartikainen et al. 2012). |
URL | http://jmlr.csail.mit.edu/papers/v15/ |
Title | Literatin |
Description | An extension that assesses the readability of text using SMOG, and cross-compared with popular and classic literature. The principle behind the plug-in is to allow users to explore the complexity of any text found within a webpage. Users identify the text of interest, activate the plug-in and the text is automatically captured and checked for readability using a SMOG based algorithm. A cross-comparison is then made between the complexity and length of the input text, with values stored for existing popular and classic literature. The idea is to present information about text complexity in a way users can relate to. |
Type Of Technology | Webtool/Application |
Year Produced | 2012 |
Impact | Literatin presents information about text complexity in a way users can relate to. |
URL | https://chrome.google.com/webstore/detail/literatin/igpbgncfceidjgcjddcdamjgbpdcmbnl |
Title | ML for mobility analysis |
Description | Understanding human mobility patterns is a significant research endeavour that has recently received considerable attention. Developing the science to describe and predict how people move from one place to another during their daily lives promises to address a wide range of societal challenges: from predicting the spread of infectious diseases, improving urban planning, to devising effective emergency response strategies. Individuals are also set to benefit from this area of research, as mobile devices will be able to analyse their mobility pattern and offer context-aware assistance and information. For example, a service could warn about travel disruptions before the user is likely to encounter them, or provide recommendations and mobile vouchers for local services that promise to be of high value to the user, based on their predicted future plans. More ambitiously, control systems for home heating and electric vehicle charging could be enhanced with knowledge of when the user will be home. |
Type Of Technology | Software |
Year Produced | 2013 |
Impact | ? Rahwan, I., Dsouza, S., Rutherford, A., Naroditskiy, V., McInerney, J., & Venanzi, M., et al. (2013). Global manhunt pushes the limits of social mobilization. IEEE Computer. Get Bibtex Citation ? McInerney, J., Rogers, A., & Jennings, N. R. (2012). Improving location prediction services for new users with probabilistic latent semantic analysis. In 4th International Workshop on Location-Based Social Networks. Get Bibtex Citation Download as PDF ? McInerney, J., Stein, S., Rogers, A., & Jennings, N. R. (2012). Exploring periods of low predictability in daily life mobility. In Mobile Data Challenge by Nokia Newcastle upon Tyne, UK: Nokia. Get Bibtex Citation |
URL | https://pypi.python.org/pypi/vbihmn/ |
Title | NILMTK |
Description | Non-Intrusive Load Monitoring (NILM) is the process of estimating the energy consumed by individual appliances given just a whole-house power meter reading. In other words, it produces an (estimated) itemised energy bill from just a single, whole-house power meter. NILM is sometimes called: •"non-intrusive appliance load monitoring (NALM or NIALM)" •"[electriciy | energy | smart meter] disaggregation" NILMTK is a toolkit designed to help researchers evaluate the accuracy of NILM algorithms. NILMTK is not yet suitable for end users because NILMTK is not yet capable of out-of-the-box disaggregation (i.e. disaggregation where you do not yet have submetered training data), although we hope that it might be some time in the future. |
Type Of Technology | Software |
Year Produced | 2014 |
Open Source License? | Yes |
Impact | NILMTK v0.2 includes a combinatorial optimisation NILM algorithm. FHMM and Hart's 1985 algorithms coming soon to v0.2. More algorithms in the pipeline! Publications 1.Nipun Batra, Jack Kelly, Oliver Parson, Haimonti Dutta, William Knottenbelt, Alex Rogers, Amarjeet Singh, Mani Srivastava. NILMTK: An Open Source Toolkit for Non-intrusive Load Monitoring. In: 5th International Conference on Future Energy Systems (ACM e-Energy), Cambridge, UK. 2014. DOI:10.1145/2602044.2602051 arXiv:1404.3878 [slides] [bib] 2.Nipun Batra, Jack Kelly, Oliver Parson, Haimonti Dutta, William Knottenbelt, Alex Rogers, Amarjeet Singh, Mani Srivastava. NILMTK: An Open Source Toolkit for Non-intrusive Load Monitoring". In: NILM Workshop, Austin, US. 2014 [pdf] [Presented as a 30-minute IPython demo of NILMTK v0.1 and a poster.] 3.Jack Kelly, Nipun Batra, Oliver Parson, Haimonti Dutta, William Knottenbelt, Alex Rogers, Amarjeet Singh, Mani Srivastava. Demo Abstract: NILMTK v0.2: A Non-intrusive Load Monitoring Toolkit for Large Scale Data Sets. In the first ACM Workshop On Embedded Systems For Energy-Efficient Buildings, 2014. DOI:10.1145/2674061.2675024. arXiv:1409.5908 [Won best demo award] [bib] [Presented as an IPython demo of v0.2 and a poster] |
URL | https://github.com/nilmtk/nilmtk |
Title | PROV Toolbox |
Description | Provenance is a record that describes the people, institutions, entities, and activities involved in producing, influencing, or delivering a piece of data or a thing. In particular, the provenance of information is crucial in deciding whether information is to be trusted, how it should be integrated with other diverse information sources, and how to give credit to its originators when reusing it. In an open and inclusive environment such as the Web, where users find information that is often contradictory or questionable, provenance can help those users to make trust judgements. PROV is a set of W3C specifications defining a model, corresponding serializations and other supporting definitions to enable the inter-operable interchange of provenance information in heterogeneous environments such as the Web. ProvToolbox is a Java library to create Java representations of PROV-DM, and convert them between RDF, PROV-XML, PROV-N, and PROV-JSON |
Type Of Technology | Software |
Year Produced | 2013 |
Open Source License? | Yes |
Impact | Wide international take up of tool, as utilising the W3C Provenance Standard |
URL | http://lucmoreau.github.io/ProvToolbox/ |
Title | Smart Home Network |
Description | The smart home framework (SHF) is a software framework aimed to simplify and speed up the modelling of smart infrastructure (i.e., smart home, smart communities and the smart grid) in the research community. It provides basic software components (e.g., appliances, generators and storage) that can be assembled together to build general models of smart homes and communities. These components can be extended to construct specialised models to meet the specific needs of any particular project. This approach enables rapid, bottom-up modelling of smart homes and communities with resuable models and components. The SHF emphasises on reusability which we believe will eventually lead to standardised models of appliances (e.g., washing machines, ovens, fridges), smart homes and smart communities that can be made available online for all. |
Type Of Technology | Webtool/Application |
Year Produced | 2013 |
Impact | In a nutshell, the framework contains the code for modelling smart homes and communities. Also, it comes with two additional toolkits for optimisation and visualisation for smart home and community data. |
URL | http://www.smarthomeframework.ecs.soton.ac.uk/ |
Title | Xively Radiation Dataset |
Description | A dataset containing nuclear radiation readings from crowdsourced Geiger counter devices installed in Japan through the Xively platform |
Type Of Technology | New/Improved Technique/Technology |
Year Produced | 2014 |
Impact | Venanzi, M., Rogers, A., & Jennings, N. R. (2013). Crowdsourcing Spatial Phenomena Using Trust-Based Heteroskedastic Gaussian Processes. In First Conference on Human Computation and Crowdsourcing (HCOMP). |
URL | http://eprints.soton.ac.uk/354861/ |
Company Name | Joulo |
Description | Joulo develops energy monitoring systems that enable customers to download their energy usage data and receive advice on ways to minimise their energy consumption. |
Year Established | 2013 |
Impact | Its brilliance is recognized by winning the 2013 British Gas Innovation Award. The technology was acquired by Quby in early 2015. |
Website | http://www.joulon.com |
Description | Autonomy and Advanced Systems Showcase |
Form Of Engagement Activity | A formal working group, expert panel or dialogue |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Policymakers/politicians |
Results and Impact | The multi-UAV controller developed by ORCHID researchers was demonstrated to David Willetts, Minister of State for Universities and Science, and Don Nutbeam, Vice Chancellor of the University of Southampton. The demo was part of the Autonomy and Advanced Systems showcase event at the University of Southampton where key representatives from industry were invited to get a flavour of the latest research outputs of the University and discuss potential tie-ups with research groups. |
Year(s) Of Engagement Activity | 2014 |
Description | BBC Summer of Wildlife 2013 |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Public/other audiences |
Results and Impact | The New Forest Cicada Project hands on activities were presented at the BBC Summer of Wildlife Event in Birmingham on 14 September 2013. There were over 4,000 visitors to this event. |
Year(s) Of Engagement Activity | 2013 |
Description | British Science Festival, Birmingham |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Schools |
Results and Impact | A software game based on a disaster response scenario for an audience of secondary school children. This sparked discussion around the use of technology in humanitarian aid. |
Year(s) Of Engagement Activity | 2015 |
Description | British Science Festival, Newcastle |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Schools |
Results and Impact | Groups of up to 30 secondary school children took part in an electronic Cicada Hunt, using mobile phones. This prompted discussion of biosound research and the use of 'smart' technology. |
Year(s) Of Engagement Activity | 2014 |
Description | FIrst Lego League Event |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | Regional |
Primary Audience | Schools |
Results and Impact | Presentation on autonomous systems and judging of the robot building competition |
Year(s) Of Engagement Activity | 2013 |
Description | I'm a Scientist |
Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Schools |
Results and Impact | For two weeks in June this year, Gopal Ramchurn participated in the national public engagement competition known as in "I'm a Scientist, Get Me Out of Here". This is an X-factor style competition where primary school students get to meet and interact with scientists. Thus, in the Cadmium Zone more than 400 students from several schools asked over 17,000 questions, with over 9000 answers being given. Gopal had the highest number of page views (over 1000 in two weeks!) and answered lots of questions on robots, apps, and technology (all relevant to ORCHID) but also questions such as 'Will AI become complex enough to become self aware?', 'why doesn't the world fall down?', 'why can't dogs and cats breed?' and 'why is mercury poisonous?'. Gopal and other scientists participated in live half-hour sessions and offline over a number of days. Every day of the final week, there were the dreaded evictions which Gopal managed to avoid till the last day! He eventually came second in his zone and thoroughly enjoyed the experience! |
Year(s) Of Engagement Activity | 2013 |
URL | http://imascientist.org.uk/ |
Description | ORCHID Industry Day(s) |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Industry/Business |
Results and Impact | A series of presentations, panel discussion and exhibitions of human-agent collective demos. This generated further interest, opportunities for internships for research students and follow up activities (including grants). |
Year(s) Of Engagement Activity | 2012,2013,2014,2015 |
Description | ORCHID Showcase |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | The event comprised an exhibition of human-agent collective demos, academic presentations, press event and interviews and a panel discussion. |
Year(s) Of Engagement Activity | 2015 |