Title:

Adaptive Selection of a Process Engine within a DBPMS

Description:

This thesis explores the design and implementation of routing functionality for the BPMN engine in PROCEED, a project with the goal to manage and execute BPMN processes in a distributed fashion.

The task was to build a router which is able to adaptively select one engine within many, optimally suited for the execution of a specific BPMN activity. To filter potential candidates, hard constraints were designed and implemented as Boolean expressions to define elimination criteria. Further a multitude of mathematical optimisation algorithms were evaluated. Of these algorithms a weighted score-based optimisation algorithm was implemented to determine one or many Pareto-optimal candidates between all engines. The final implementation was benchmarked quantitatively and against a set of qualitative requirements for machines designed for the Internet of Things.

Supervisor: Kai Grunert

Type:  Bachelor Thesis

Duration: 5 months