Studentische Arbeiten

Diplomarbeiten, Masterarbeiten, Bachelorarbeiten, Studienarbeiten, Fachstudien

Student Projects

The department is currently not able to accept any further projects, due to over-saturation. This page will be updated as the situation changes. 

The following projects can be adapted for Diploma Theses, Master Theses, Bachelor Theses, Research Projects (Studienarbeiten) and Student Reports (Fachstudien). Contact Prof. Dr. Marco Aiello in case you have further questions or would like to submit your own proposals.

For Bachelor and Master thesis projects, please follow these instructions:

  1. Use the registration form in Campus (including the achieved credit points)
  2. Fill in the contract and the licence agreement on-line including your private email in the license agreement. 
  3. Bring signed (original) contract, licence agreement, anouncement, and the registration to Ms. Ibach
  4. Bring the registration form to the Examination office in order to register the thesis and one copy to Ms. Ibach for archiving
  5. Please notice that there is a hard deadline of 6 months from the date of registration. The date of registration cannot be on a weekend or official German holiday.

Note: Concepts and implementation produced in any project must be compatible and published under the MIT license.


 

Demand-side management via AI planning

Commercial buildings are characterised by high operational costs and high CO2 emission levels due to the high demand for energy [1]. The management of energy demand, also known as demand-side management, allows buildings to make informed decisions about their energy consumption, but it also helps energy providers to shape load demand profiles [2]. However, current building management systems are capable of controlling heating, ventilation and air conditioning, lighting and other aspects using specific strategies limited by basic scheduling functions [3]. On the other hand, with the proliferation of the Internet of Things (IoT) deployed in building, building management systems can be transformed into systems employing intelligent demand-side management. In other words, buildings equipped with various sensors and actuators have the potential to become intelligent and provide occupants with a more comfortable experience, reduced operational costs, and improved energy efficiency.

This calls for developing approaches that can plan the energy demand of buildings intelligently. In this context, Artificial Intelligence (AI) planning offers powerful techniques for automated and dynamic selection and organisation of actions that, when executed, achieve a given objective [4]. AI planning could be helpful for buildings to intelligently plan their demand by computing effective plans or schedules of device actions.

The aim of this project is to propose and develop an approach for demand-side management via AI planning. This involves defining the problem of demand-side management as a domain-independent planning problem, modelling the planning problem using a standard syntax, such as PDDL [5], solving the planning problem using existing AI planners, and evaluating the efficiency and effectiveness of the proposed approach.

Keywords: AI planning, demand-side management, Internet of Things, office buildings

Literature

[1] D. D'Agostino, B. Cuniberti, and P. Bertoldi, Energy consumption and efficiency technology measures in European non-residential buildings," Energy and Buildings, vol. 153, pp. 2-86,2017.

[2] P. Palensky and D. Dietrich, Demand side management: Demand response, intelligent energy systems, and smart loads," IEEE Transactions on Industrial Informatics, vol. 7, no. 3, pp. 381-388, 2011.

[3] I. Georgievski, T. A. Nguyen, F. Nizamic, B. Setz, A. Lazovik, and M. Aiello, Planning meets activity recognition: Service coordination for intelligent buildings," Pervasive and Mobile Computing, vol. 38, no. 1, pp. 110-139, 2017.

[4] M. Ghallab, D. S. Nau, and P. Traverso, Automated planning: Theory & practice. Morgan Kaufmann Publishers Inc., 2004.

[5] M. Fox and D. Long, PDDL2.1: An extension to PDDL for expressing temporal planning domains," Journal of Artificial Intelligence Research, vol. 20, no. 1, pp. 61-124, 2003.

Contact: Dr. Ilche Georgievski


 

Learning planning domains for office buildings using IoT data

Commercial buildings, such as office buildings, are characterised by high operational costs, high CO2 emission levels due to the high energy demand, and limited comfortability [1]. Current building management systems control building operations in terms of heating, ventilation, air conditioning and lighting using specific strategies limited by basic scheduling functions [2]. On the other hand, with the proliferation of the Internet of Things (IoT) deployed in buildings, such systems can be transformed into entities capable of intelligent decision-making. In other words, buildings equipped with various sensors and actuators have the potential to become intelligent and provide occupants with an improved experience in terms of comfort and productivity, reduced operational costs, and improved energy efficiency.

This calls for employing approaches that can plan the energy demand of buildings intelligently. In this context, Artificial Intelligence (AI) planning offers powerful techniques for automated and dynamic selection and organisation of actions that, when executed, achieve a given objective [3]. AI planning could be helpful for buildings to intelligently plan their operation by computing effective plans or schedules of device actions, but also to assist their occupants by computing plans of user actions that can improve their comfort and productivity. To compute such plans of action, AI planning requires a planning domain model in which actions need to be specified in terms of preconditions that need to hold so that an action can be applied and postconditions as effects that the action has after its application. Planning domains are typically authored manually by domain experts, which is an arduous and error-prone process. Alternatively, such planning domains can be automatically generated using machine learning techniques [4].

The aim of this project is to develop a system that can learn planning domain models for office buildings using IoT data. The main project tasks include:

  • Examine the current state of the art on learning planning domains for AI planning with an emphasis on planning domains for commercial buildings.
  • Develop an approach for learning planning domains from IoT data. This requires analysis and selection of appropriate machine learning techniques.
  • Use existing IoT datasets and/or deploy IoT in an environment to gather the necessary data.
  • Learn planning domains specified in the Planning Domain Definition Language (PDDL) [5].
  • Evaluate the approach in terms of the effectiveness of learning PDDL planning domains. Also, evaluate the suitability of the learned PDDL planning domain(s) when used by existing planners, such as FF [6] and FD [7], to compute plans.

Keywords: AI planning, machine learning, Internet of Things, office buildings

Literature

[1] D. D'Agostino, B. Cuniberti, and P. Bertoldi, Energy consumption and efficiency technology measures in European non-residential buildings," Energy and Buildings, vol. 153, pp. 2-86, 2017.

[2] I. Georgievski, T. A. Nguyen, F. Nizamic, B. Setz, A. Lazovik, and M. Aiello, Planning meets activity recognition: Service coordination for intelligent buildings," Pervasive and Mobile Computing, vol. 38, no. 1, pp. 110-139, 2017.

[3] M. Ghallab, D. S. Nau, and P. Traverso, Automated planning: Theory & practice. Morgan Kaufmann Publishers Inc., 2004.

[4] S. Jimenez, T. De La Rosa, S. Fernandez, F. Fernandez, and D. Borrajo, A review of machine learning for automated planning," The Knowledge Engineering Review, vol. 27, no. 4, pp. 433-467, 2012.

[5] M. Fox and D. Long, PDDL2.1: An extension to PDDL for expressing temporal planning domains," Journal of Artificial Intelligence Research, vol. 20, no. 1, pp. 61-124, 2003.

[6] J. Hoffmann. (1997) Fast forward homepage. [Online]. Available: https://fai.cs.uni-saarland.de/hoffmann/ff.html

[7] F. Downward. (2003) Fast downward homepage. [Online]. Available: http://www.fast-downward.org/

Contact: Dr. Ilche Georgievski


 

Who cites whom and who cites what in AI planning

Artificial Intelligence (AI) planning provides powerful techniques for searching and selecting actions that, when executed, achieve a given goal [1]. The field of AI planning is a long-standing one and the amount of research is so big and increasing rapidly that understanding exactly how far the field has progressed has become more difficult. On the other hand, the interest and scope for exploitation of AI planning in industry and commerce are continuously increasing. For example, automation is a trend that requires efficient and robust AI planning. So, it is significant for both practitioners and scientists to understand the current state of the field of AI planning and especially the emerging research topics among the numerous publications. Since providing a systematic overview of the field manually by experts is arduous and time-consuming, a computation approach seems more appropriate. Complex network analysis offers quantitative techniques for studying the structures and characteristics of networked systems [2]. Citation network analysis is especially relevant because it focuses on exploring the frequency and patterns of citations in documents, including academic publications [3]. It has been used for the analysis of various fields, such as learning analytics [4], regenerative medicine [5], organic photovoltaic cells [6], human-resource development [7], and others.

The aim of this project to evaluate the current state of the field of AI planning by (citation) network analysis. The focus is on articles appearing in the ICAPS, IJCAI and AAAI conferences. The questions of interest are:

  • What structural attributes and characteristics are present in the network(s) of AI planning? Are there any specificities in knowledge flows between countries, institutions, and technology fields in the field of AI planning?
  • What are the main interests, topics, and issues of AI planning? Which trends are emerging? Which disciplinary hierarchies have been influencing the development of the field so far? Which are under-represented areas of research? Which disciplines may require more strategic and targeted support and funding opportunities? Are cited papers of a more conceptual or empirical nature?
  • Which authors and publications are most influential to the field of AI planning, and to which specific areas of automated planning do they contribute knowledge?
  • Are there distinctive subgroups in the network(s) of AI planning? If yes, what are they? Are there in-group citations?

Keywords: complex network analysis, citation network, AI planning

Literature

[1] M. Ghallab, D. S. Nau, and P. Traverso, Automated planning: Theory & practice. Morgan Kaufmann Publishers Inc., 2004.

[2] M. Newman, Networks: An Introduction. USA: Oxford University Press, Inc., 2010.

[3] H. Small, Tracking and predicting growth areas in science," Scientometrics, vol. 68, no. 3, pp. 595-610, 2006.

[4] S. Dawson, D. Gasevic, G. Siemens, and S. Joksimovic, Current state and future trends: A citation network analysis of the learning analytics field," in International Conference on Learning Analytics And Knowledge, ser. LAK '14, 2014, pp. 231-240.

[5] N. Shibata, Y. Kajikawa, Y. Takeda, I. Sakata, and K. Matsushima, Detecting emerging research fronts in regenerative medicine by the citation network analysis of scientific publications," Technological Forecasting and Social Change, vol. 78, no. 2, pp. 274-282, 2011.

[6] H. Choe, D. H. Lee, I. W. Seo, and H. D. Kim, Patent citation network analysis for the domain of organic photovoltaic cells: Country, institution, and technology field," Renewable and Sustainable Energy Reviews, vol. 26, no. 2013, pp. 492-505, 2013.

[7] S. J. Jo, C.-W. Jeung, S. Park, and H. J. Yoon, Who is citing whom: Citation network analysis among HRD publications from 1990 to 2007," Human Resource Development Quarterly, vol. 20, no. 4, pp. 503-537, 2009.

Contact: Dr. Ilche Georgievski


 

Development of an application to calculate hourly CO2-emission intensity of European countries

More than 40% of global carbon emissions are caused by electricity generation. The CO2 emissions tied to the consumption of electricity depend on how this is generated. For instance, coal-fired power plants emit 950 grams of CO2 for every kilowatt-hour of electricity they generate, while gas-fired and nuclear power plants emit roughly 350 gCO2/kWh and 12 gCO2/kWh,respectively. Assessing the environmental impact of electricity consumption is usually based on the average CO2-emission intensity of the generation mix, including energy import and export.

The goal of this study project is to develop an application to acquire and manipulate hourly data about the generation and trade of electricity in European countries, using the information available on the ENTSO-E Transparency Platform [1]. Based on the flow tracing method [2], the application will calculate the hourly average CO2-emission intensity for each country [3, 4]. Moreover, it will allow for statistical analyses based on common metrics, e.g., cross- and autocorrelation.

The application will be developed in Python.

Concepts and implementation produced in this project must be compatible and published under the MIT license.

Literature

[1] ENTSO-E, "ENTSO-E Transparency Platform," jan 2020. [Online]. Available: https://transparency.entsoe.eu/

[2] J. Hörsch, M. Schäfer, S. Becker, S. Schramm, and M. Greiner, "Flow tracing as a tool set for the analysis of networked large-scale renewable electricity systems," International Journal of Electrical Power and Energy Systems, vol. 96, no. April 2017, pp. 390-397, 2018.

[3] B. Tranberg, O. Corradi, B. Lajoie, T. Gibon, I. Staffell, and G. B. Andresen, "Real-time carbon accounting method for the European electricity arkets," Energy Strategy Reviews, vol. 26, 2019.

[4] "Real-time carbon accounting method for the European electricity markets," 2019. [Online]. Available: https://arxiv.org/abs/1812.06679 

Contact: Laura Fiorini, M. Sc.Prof. Dr. Marco Aiello


 

Designing and implementing the capabilities of AI planning as services

Artificial Intelligence (AI) planning provides powerful techniques for searching and selecting actions that, when executed, achieve a given goal [1]. Even though AI planning has made significant progress in terms of advanced algorithms and representations, it hardly takes part in real-world applications. To integrate planning in an application, a developer must get bogged down in theory and detail. This is because most AI planning systems are not aware of any
software design principles.

A common way to transparently accomplish interoperability, reusability, distribution, and reduction of information technology burden is by considering the key elements of Service-Oriented Computing (SOC) [2, 3]: service orientation, services, service composition, and service-oriented architecture. Among these key elements, service orientation provides a set of design principles to be used as guidelines for realising specific but common characteristics of the applications of interest [4].

The aim of this project is to explore the application of service orientation in AI planning, find a classification of planning services depending on their tasks and entities they use, design and implement a set of planning services, and demonstrate the interoperability, flexibility in terms of arbitrary application compositions, efficiency in terms of rapid application prototyping, and the potential of reuse of such designed planning services.

Literature

[1] M. Ghallab, D. S. Nau, and P. Traverso, Automated planning: Theory & practice. Morgan Kaufmann Publishers Inc., 2004.

[2] M. P. Papazoglou and D. Georgakopoulos, “Introduction: Service-oriented computing,” Commun. ACM, vol. 46, no. 10, pp. 24–28, 2003.

[3] T. Erl, SOA principles of service design. Prentice Hall PTR, 2007.

[4] I. Georgievski, “Coordinating services embedded everywhere via hierarchical planning,” Ph.D. dissertation, University of Groningen, October 2015.

Contact: Dr. Ilche Georgievski


 

CO2-efficient demand response: potential benefits in varying energy mix scenarios

Due to high energy consumptions, residential and office buildings are responsible for one third of total CO2 emissions in the European Union (European Commission).  The need for electricity has raised significantly due to the increasing ownership of appliances (Haines, 2010), and, as a consequence, its environmental impact, which varies according to weather conditions, time of the day, production sources, and location. Traditional demand-response programs aim to control the residential electric use in response to price signals or incentive payments (Siano, 2014). Another possible approach is to schedule generation and consumptions according to CO2 signals (Fiorini & Aiello, 2019). Yet, the potential benefits of such programs depend on how the electricity is generated  (Claire Bergaentzlé, 2014; Stoll, 2014).

The research proposed here is to investigate to what extent CO2-efficient demand response programs could enable emission reductions in various countries. The comparison should include countries with different energy generation parks, e.g., France with high penetration of nuclear, Germany with high penetration of coal and wind, The Netherlands with high penetration of natural gas. A quantitative assessment of the potential benefits should also be estimated in varying energy mix scenarios.

Literature

Claire Bergaentzlé, C. C. (2014). Demand-side management and European environmental and energy goals: An optimal complementary approach. Energy Policy, 858-869.

ENTSO-E. (n.d.). Transparency Platform. Retrieved from https://transparency.entsoe.eu/

European Commission. (n.d.). Buildings. Retrieved from https://ec.europa.eu/energy/en/topics/energy-efficiency/buildings

Fiorini, L., & Aiello, M. (2019). Predictive CO2-efficient scheduling for hybrid electric and thermal loads. IEEE International Conference on Energy Internet, (pp. 392-397). Nanjing.

Haines, V. L. (2010). How Trends in Appliances Affect Domestic CO2 Emissions: A Review of Home and Garden Appliances. Department of Energy and Climate Change.

Siano, P. (2014). Demand Responde and Smart Grids--A Survey. Renewable and Sustianable Energy Reviews, 461--478.

Stoll, P. B. (2014). Including dynamic CO2 intensity with demand response. Energy Policy, 490-500.

Contact: Laura Fiorini, M. Sc.Prof. Dr. Marco Aiello


 

The Integration of Electric Vehicles in the Smart Grid

Electric Vehicles (eV) have a battery size that can vary from few KWh to almost 100 KWh. These are frequently connected to the power grid to charge. Several researchers have proposed the idea that these batteries can also be used to store grid electricity and provided it back to grid during peak demand periods. The research proposed here is to assess technical solutions to make the eV an integral part of the power grid, to study individual strategies based on dynamic pricing for incentivising eV integration, and to assess the economic viability of such approaches.

Literature

A. P. Lopes, F. J. Soares and P. M. R. Almeida, "Integration of Electric Vehicles in the Electric Power System," in Proceedings of the IEEE, vol. 99, no. 1, pp. 168-183, Jan. 2011.

Francis Mwasilu, Jackson John Justo, Eun-Kyung Kim, Ton Duc Do, Jin-Woo Jung, “Electric vehicles and smart grid interaction: A review on vehicle to grid and renewable energy sources integration,” in Renewable and Sustainable Energy Reviews, vol. 34, no. 6, pp 501-516, Jun. 2014.

Contact: Prof. Dr. Marco Aiello


 

Service Composition for Personal Assistants

Personal Assistants like Siri, Viv, Alex, Cortana, rely on online services to satisfy the requests of a user. Most often a request cannot be satisfied with just one service, but it needs the composition of multiple services. In this project you will investigate techniques to intelligently compose services  and you will implement prototypes connecting to Web services to satisfy complex user requests. 

Literature

Kaldeli, A. Lazovik, and M. Aiello (2016) Domain-Independent Planning for Services in Uncertain and Dynamic Environments, Artificial Intelligence, Elsevier, 236(7): 30-64.

Contact: Prof. Dr. Marco Aiello


 

Modelling and Simulating Research Funding Distribution in Computer Science

Today researchers heavily rely on research grants and subsidies to finance their research. The attribution of such subsidies is usually based on a peer-evaluation process that awards grants to the most promising proposals. In this project you will perform a statistical analysis of grant acquisition. You will provide a model to predict grant distribution and simulate various scenarios to identify the effects of specific policies on the probabilistic distribution of grants and on their impact on research and innovation. Bibliometric data for complex network analysis will be part of the developed model.

Contact: Prof. Dr. Marco Aiello


 

Cellular Automata for Modelling Resource Allocation

Cellular automata are simple models that can represent complex even chaotic systems. The goal of this project is to provide simple models for resource allocation that help identifying efficient allocation rules. The project entails both a modelling and simulation component to it.

Contact: Prof. Dr. Marco Aiello


 

A Complex Network Analysis of Models to be Checked

Model Checking is the field concerned with the automatic process of checking whether a given model (or family of models) meets certain specifications. Model checkers work in various ways usually exploring large models explicitly represented as states or symbolically represented. Your task is to evaluate the impact of using complex network analysis techniques to improve and extend model checking approaches.

Contact: Prof. Dr. Marco Aiello


 

Kontakt

Marco Aiello
Prof. Dr.

Marco Aiello

Abteilungsleiter

Elisabeth Ibach
 

Elisabeth Ibach

Sekretariat

Ilche Georgievski
Dr.

Ilche Georgievski

Leitung Arbeitsbereich Smart Energy Systems

Zum Seitenanfang