Smart Cities / Smart Grids

Decision Support Systems to Manage Flexible Workflows

Business process management has become a crucial issue for many industrial processes. The accurate definition of the steps to be taken and of the resources needed in each process is important to manage the evolution and the interactions of the business activities. Workflows are useful tools to describe the sequence of execution of the activities present in business processes and their relationships. As several workflows may be underway at the same time, using and sharing resources, an efficient workflow management system is essential to support process managers take the right decisions at the right times. That is why several experienced researchers in the fields of decision support systems and resource optimization have gathered to face the challenges posed by workflow management. The project will result in new methods to detect instances of activity workflows underway, to monitor workflows using limited and shared resources, to detect deviations from the expected execution and to take corrective measures if necessary, and to discover new workflows. The fundamental research conducted to achieve the goals is based on the following techniques:

• Data mining methods to learn about workflows
• Case-based reasoning for decision support systems
• Robust and egalitarian optimization methods to assign resources in distributed workflows

As far as applications are concerned, the project will apply the results to four experimental scenarios the group has been involved with for several years. In this way the results will be benchmarked against previous results using previous techniques and the benefits of the new methods will be computed.