EPS-AIST 2017 Abstracts

Area 1 - Software Agents and Internet Computing
Full Papers
Full Papers
Paper Nr: 1
Title:

NATCONSUMERS - NATural Language Energy for Promoting CONSUMER Sustainable Behaviour
Authors:

Zoltan Kmetty, Sándor Molnár and Manolisz Karajánnisz
Abstract: Residential energy consumption represents 28% of all EU energy consumption, and if commercial buildings are also considered this percentage increases to 40% (36% of EU CO2 emissions). In this context, it is clear that the reduction of consumption in the residential sector should play an important role in energy efficiency programs and policies, as is stated in the Energy Efficiency Directive 2012/27/EU. Most energy efficiency measures implemented in Europe involved technological interventions. In contrast, every-day energy-consuming behaviours are largely habitual and therefore the potential for energy savings at home with actions focused on consumer behaviour is very promising. NATCONSUMERS is a project designed to improve feedback to consumers about energy usage within the European residential sector. The project is developing a methodology for communicating more effectively with consumers by using ‘Natural Language’. The NATCONSUMERS tool will be designed to raise awareness about how people use energy within their homes and to give them advice about how to use energy more sustainably. In order to send tailored messages, multiple segmentations based on load profiles (from smart meters) and attitudinal surveys were made. Based on user load profiles we search for typical load curves, which could distinguish different types of user profiles. In order to compare the household’s energy consumption to benchmarks we segmented the volume of consumption by demographics with various statistical models. The result of load profile and demographic segmentation defines the content of the message send to the users. We are convinced that communication efficiency depends not only on the content of the message (“What”), but is also very much influenced by the messaging style and argumentation (“How”). Therefore different segmentations are needed, including load profile classification and segmentation based on demographics, as well as segmentation based on social values and attitudes. Building on these segmentation results, we have defined an energy feedback framework based on the provision of personalised recommendations in natural language tailored to each consumer group. The recommendations compare the historical consumption of the user with households showing similar characteristics, and provide possible measures to address that takes into account their personal situation and context. Tailored energy advice has been found, through multiple studies, to be much more effective than provision of more general advice. Variations in consumers’ characteristics and behaviours lead to heterogeneous energy demands, influenced by both individual preferences and physical variables. Furthermore, framing advice in differing ways for different consumer can improve the likelihood of the advice being acted upon. With the help of NATCONSUMERS framework we expect a 5-10 percent electricity reduction could be achieved in the residential segment. The NATCONSUMERS framework could be adopted by any third party user who deals with energy consumers and has access to their Smart-Meter data. The developed framework provides a complete description of an advice tool that any energy utility or any other kind of third party user (like consumer associations) could apply in their work with consumers. This tool would help to reduce the distance between energy utilities and consumers (i.e. minimize trust issues), and also help to raise awareness of electricity usage. The positive change of consumers’ behaviour with the reduction of their electricity usage also helps to decrease CO2 emissions.
Paper Nr: 2
Title:

CloudSocket
Authors:

Frank Griesinger, Robert Woitsch, Kyriakos Kritikos, Daniel Seybold and Jörg Domaschka
Abstract: The project aims to encompass a "hybrid process" modelling framework that reconciliates semantic inference, rule-based inference, meta-modelling management techniques and knowledge management techniques in order to bring SMEs closer to the Cloud by making it attractive for them to incorporate cloud resources and components for their realization of their goals. The proposed framework implements a layered approach for managing the complexity of bridging the semantic distance from business process to workflow configuration of Business Processes in the Cloud. (1) A Knowledge-Based Approach: For bridging the gap between the business and IT level and enabling the presentation of monitoring results and other type of evaluation suggestions at the business level. (2) Business Processes in the Cloud: To deal with their management and deployment through defining and executing self-adaptive workflows in multi-cloud environments. The management of business processes, by including the monitoring and evaluation of KPIs and SLOs, also involves their respective adaptation via executing adaptation rules to sustain the service levels promised to the business process customers. (3) Substanstial SME support: CloudSocket aims to provide business solutions to the SME market which can be offered in an open and interoperable way. A particular focus is on startups which desire not to invest in their own infrastructure but concentrate mainly on developing their business. Thus, as a business is modified and progressed in the course of time, the respective IT services have to be flexibly adapted and this will be definitely supported by CloudSocket.
Paper Nr: 3
Title:

CloudLightning

Authors:

Anna Gourinovitch, David Kenny, Theo Lynn, Huanhuan Xiong, Anne Elster, Malik Khan, Konstantinos Giannoutakis, Ioan Dragan, Dana Petcu, Dapeng Dong, John Morrison, George A. Gravvanis, Perumal Kuppuudaiyar, Christos Papadopoulos - Filelis, Marian Neagul and Suryanarayanan Natarajan
Abstract: CloudLightning is a three-year research project funded by the European Commission’s Horizon 2020 programme for research and innovation. It intends to develop an intelligent, power-efficient cloud computing infrastructure that will provide energy savings to cloud service providers and simplify access to cloud resources for cloud consumers. The project’s vision of a heterogeneous self-organising cloud assumes a cloud that makes use of different specialist hardware devices that can accelerate the completion of specific tasks or can be put in a state where less power is used (or indeed deactivated if possible) when not required, thus maximising both performance and energy efficiency. The availability and utilisation of heterogeneous resources in the cloud will open up the cloud to a wider range of stakeholders, particularly in the compute-intensive application space: (1) Genome Processing, (2) Oil and Gas exploration, and (3) Ray Tracing (3D image rendering). Expected impacts for cloud service providers that adopt the CloudLightning delivery model include increased competitiveness through reduced cost and differentiation, increased energy efficiency and reduced environmental impact, improved service delivery, and greater accessibility to cloud computing for high performance computing workloads.
Paper Nr: 4
Title:

RECAP - Reliable Capacity Provisioning and Enhanced Remediation for Distributed Cloud Applications
Authors:

Anna Gourinovitch, Theo Lynn, Jose Lopez Veiga, Stefan Wesner, P-O Östberg, Jörg Domaschka, James Byrne, Miguel Angel Lopez Pena, Sergej Svorobej, John Kennedy, Vincenzo Mancuso, Philip Eardley, Johan Forsman, Radhika Loomba, Paolo Casari, Anders Torneus, Manuel Noya Marino, Peter Willis and Antonio Fernandez Anta
Abstract: The REliable CApacity Provisioning and enhanced remediation for distributed cloud applications (RECAP) project aims to advance cloud and edge computing technology, to develop mechanisms for reliable capacity provisioning, and to make application placement, infrastructure management, and capacity provisioning autonomous, predictable and optimized. Large-scale computing systems are today built as distributed systems where components and services are distributed and accessed remotely through clients and devices. However, while recent years have seen significant advances in system instrumentation as well as data centre energy efficiency and automation, computational resources and network capacity are often provisioned using best effort provisioning models and coarse-grained quality of service (QoS) mechanisms. RECAP will go beyond the current state of the art and develop the next generation of cloud/edge/fog computing capacity provisioning via targeted research advances in cloud infrastructure optimization, simulation and automation. The overarching result of RECAP is the next generation of agile and optimized cloud computing systems. The outcomes of the project will pave the way for a radically novel concept in the provision of cloud services, where services are instantiated and provisioned close to the users that actually need them by self-configurable cloud computing systems.
Paper Nr: 7
Title:

European residential demand response; a report from the SEMIAH project
Authors:

Rune Hylsberg Jacobsen
Abstract: With the advent of smart grids, new solutions for energy management become available. During the last decade, manufacturers have focused on the development of smart appliances. However, a large market uptake of smart appliances is not expected to occur in the short-term. Demand response, which is defined as changes in electricity usage by consumers from their normal consumption patterns in response to signals from the grid operations or the energy markets, is considered one of the key solutions to improve energy efficiency and for reducing peak demand. However, no automated demand response programs have been implemented for European households despite the fact that households represented approximately 27% of the total energy consumption in Europe in 2010 and were responsible for 10% of the carbon dioxide emissions in 2007. The consortium behind the Scalable Energy Management Infrastructure for Aggregation of Households (SEMIAH) project aims to pursue a major technological, scientific and commercial breakthrough by developing a novel ICT infrastructure for the implementation of demand response in households. The SEMIAH infrastructure enables the shifting of energy consumption from high energy-consuming loads to off-peak periods with a high generation of electricity from renewable energy sources. The demand response system of SEMIAH connects the Distribution System Operator (DSO) and the Transmission System Operator (TSO) over a Wide Area Network (WAN) infrastructure. This enables actors in telecommunication and the energy sector in a joint effort to pursue a more secure and sustainable energy supply for the future. In order to provide a demand response of significance to the grid operation, a lot of households must be aggregated. The project will use a simulation approach that promise to scale to at least 200.000 households and that provides a tool for an aggregator to plan their demand response programs and act accordingly on the energy markets. The simulation tool will be validated against two demonstration sites with a total of 200 households from Norway and Switzerland, respectively. The presentation will give an overview of the SEMIAH information and communication technology infrastructure solution and give insights into results from the SEMIAH pilot. In addition, the presentation will look into scalability aspects of the proposed solution.
Paper Nr: 8
Title:

Increased data reuse in Europe: An opportunity and a challenge at the same time
Authors:

Klara Süveges-Heilingbrunner
Abstract: Enormous amounts of data are available around us. Data is collected by different means and from different sources, and, so far, primarily used by the actor that collected it and for the originally intended purpose. This is good, particularly if personal data is concerned, but thereby, a significant part of the potential that lays in most data remains unexploited. Promoting and facilitating the reuse of data is considered a key to exploit the full potential. Increased reuse of data is expected to boost the European (data) economy. However, the reuse of data also raises plenty of questions, which usually need to be addressed on a European level. The general data protection regulation, the free flow of data initiative, standardization efforts, for instance, related to interoperability and awareness raising initiatives are only a few examples of answers. The process of finding appropriate answers that do not only allow overcoming the challenges but also exploiting the opportunities of increased data reuse in Europe to the greatest possible extent, would be supported by the IoTBDS 2017 activities proposed by EuDEco. The key aim is to involve the participants of the conference in a joint reflection and discussion on the current status and possible development paths of the European data economy placing special attention on data reuse. Legal, socio-economic and technological issues would be at the focus within the scope of the workshop and at the stand. Participants from academic institutions, companies and government bodies would be invited to share their experiences related to data reuse and to give feedback on what others report. The presence of EuDEco at the IoTBDS conference would contribute to the finalisation of a model of the European data economy that EuDEco is currently developing as a tool for experts to better understand how the full potential of data can be exploited. In addition, the joint reflection and discussion would contribute to the formulation of a set of concrete and actionable recommendations for both policy makers, who set the framework conditions for the European data economy and leaders of companies, government bodies and academic institutions participating in the data economy.

Area 1 - Software Agents and Internet Computing

Full Papers
Paper Nr: 6
Title:

A Parallel Approach to Data Analysis for Energy Efficiency Problem Solution
Authors:

M. Carmen Pegalajar, Manuel Isidoro Capel-Tuñón and Luis Gonzaga Baca Ruiz
Abstract: Energy efficiency advancement is considered as a research area of paramount interest by governmental institutions recently. The development of new sensor-based technologies and industrial processes has brought about monitor devices installation in a plethora of scenarios (distributed facilities [1], smart city [2]) nowadays, and therefore data availability from different aspects of life in real-time becomes a major necessity. Our final aim in this research work is to get the comprehension about how and when energy is consumed in different environments [3,4]; and therefore, we propose the development of methods and software tools for carrying out management and analysis (Intelligent Data Analysis, Data Mining Techniques, Knowledge Extraction, Temporal Series and Optimization Algorithms) of Information originated from different sources. It is widely acknowledged that the deployment and use of the aforementioned tools and methods, respectively, requires a great amount of computational power. Indeed, the models training time is specially considered one of the most expensive process when it comes to solve energy efficiency problems by computational methods and that process needs to iterate the model-training phase several times for finding optimal solutions. This need made us to explore new ways for achieving reduction of execution time by parallelization of the methods that we use for data treatment in energy efficiency problem solving. We are currently using different neural networks models for energy consumption forecasting inside Granada University buildings. In spite of the results obtained with classical neural-based methods are optimal as of today, the computation time can however become too big when solving a problem that requires obtaining good solutions in real-time. We observed that the last parallel designs of optimization models used yielded programs whose execution time was cut down by the 50% without affecting result soundness. The latter results encourage us to continue investigating the application of parallel and real-time programming to speedup different algorithms connected with our main research theme. The main subject of our presentation will be the Spanish Ministry of Economy research grant: "DIFFERENTIAL: INTELLIGENT DATA ANALYSIS FOR ENERGY EFFICIENCY MANAGEMENT IN DISTRIBUTED FACILITIES"(ANALISIS INTELIGENTE DE DATOS PARA GESTION DE LA EFICIENCIA ENERGETICA EN INSTALACIONES DISTRIBUIDAS) TIN201564776-C3-1-R Funding institution: Ministry of Economy and Competiveness Research program (funding): RETOS (Programa Estatal de I+D+i orientada a los Retos de la Sociedad) Institution type: Gubernamental Project carried out in the city of: GRANADA, Andalucía, España Principal researchers names (IP, Co-IP,...): MIGUEL DELGADO CALVO-FLORES; MARÍA DEL CARMEN PEGALAJAR JIMENEZ Starting date: 01/01/2016 Ending date: 31/12/2018 Total amount: 165.770 € The DIFFERENTIAL project can be considered a follow up of the FP7 project "ENERGY IN TIME:Simulation based control for Energy Efficiency building operation and maintenance" in which the research group of prof. M.Carmen Pegalajar Jiménez also took part as during the complete lifetime of that project.