Exjobbsförslag från företag

Detta är ett uppsatsförslag hämtat från Nationella Exjobb-poolen. Klicka här för att komma tillbaka till samtliga exjobbsförslag.

Förslaget inkom 2006-12-14

Deriving Quality of Service Agreements based on Business Process Key Performance Indicators

Service Oriented Architectures (SOA) is a new architecture paradigm for enterprise applications. Software built according to the SOA is based on services. In addition to technical concepts, the SOA reaches out to Business Process Management (BPM). Based on business process activities, services are designed to offer operations that directly support these activities. By orchestrating services in a way that is structurally equivalent to the business processes, a structural equivalence of business process to the IT systems supporting these processes can be achieved.

BPM uses Key Performance Indicators (KPI) to manage processes, defining measurable attributes for processes and evaluating the processes according to the attributes. Similarly, services in a SOA have certain quality attributes like availability or failure rate. The service quality attributes are used to manage the services. If services are used to support business processes, there presumably is a link between the values of the quality attributes and the values of the KPI.

This master thesis proposal entails developing a methodology for deriving requirements on the service quality levels for quality attributes, e.g. availability or performance, based on business process KPIs. A business process KPI could for instance be productivity or lead-time. In order to assess the impact of different values of quality attributes, the project will begin by a literature study to summarize state of the art in the scientific community regarding assessments of the business impact of information systems. This will be formally expressed in an Extended Influence Diagram (EID). The next step is to create models of business processes and their relation to the services. The last step is to simulate business process behaviour using the business process models and different values of quality of attributes of the services (e.g. varying levels of availability and it’s impact on the business process lead time) and measuring the impact on KPI.

The goals of the master thesis are
• To summarize scientific state-of-the art studies on the impact of quality attributes on business process KPIs
• To create an Extended Influence Diagram for assessing these impacts
• To develop modelling constructs (meta-models) of business process models and their relations to services based on existing meta-models (e.g. BPMN=
• To model business processes and their relations to services
• To simulate business process KPIs with varying values of quality attributes of the services, thus verifying or falsifying the theory of the EID.
• To write an article based on the master thesis.

This master thesis work will partly take place at KTH, Stockholm, and partly at T-Labs, TUB in Berlin, Germany. The student does not need to speak German, but must be fluent in English.

One goal of the master thesis project is that it will lead to a publication. Therefore, the student should also be willing to write an article together with T-Labs based on the master thesis when it’s completed. This would be compensated by the course “Individual Project in Industrial Information and Control Systems” (2C1530) which would generate 5 credits (7,5 ECTS credits).

Appropriate applicants are students from D, E, F, IT and IndEk. Documented experience and interest of IT-management and BPM or SOA are extra qualifying. Send CV and grade transcript accompanied by a personal letter to Pia Gustafsson, KTH before November 17 2006. With further questions, please contact Pia Gustafsson.

Supervisor at KTH:
Pia Gustafsson piag@ics.kth.se

Supervisor at T-labs:
Maximilian Ahrens maximilian.ahrens@telekom.de

Lodging and travel expenses to and from Berlin will be covered by T-labs.
Start-date: November 2006


Informationen om uppsatsförslag är hämtad från Nationella Exjobb-poolen.