Business Cockpit

Overview

This project aims at developing a set of integrated tools that supports business and IT users in managing process execution quality. We refer to this set of tools as the Business Cockpit (BC) tool suite. BC allows definition, monitoring and analysis of any kind of business-level metrics on top of any data source. For example, Business Process Management Systems (BPMSs) record many types of events that occur during process executions, including the start and completion time of each activity, its input and output data, the resource that executed it, and any failure that occurred during activity or process execution. By cleaning and aggregating process logs into a warehouse and by analyzing them with business intelligence technologies, we can extract knowledge about the circumstances in which high- or low-quality executions occurred in the past, and use this information to explain why they occurred as well as predict potential problems in running processes. We use a very high-level and user-oriented notion of quality: we assume that it is up to the (business or IT) users to define what quality means to them, and in general which are the characteristics that they want to analyze and predict.

Problems Addressed

The BC suite provides several features to automatically or semi-automatically manage quality:

Analysis: BC allows users to analyze completed process executions from both a business and an IT perspective. IT analysts will be interested in viewing detailed, low-level information such as average execution time per node or the length of the work queues of resources. Business users will instead be interested in higher-level information, such as the number of "successful" process executions, or the characteristics of processes that did not meet the Service Level Agreement (SLA) stipulated with customers. Besides providing a wide set of reporting functionalities, BC also offers several features to help analysts identify the causes of process execution behaviors of interest.

Prediction: BC can derive prediction and correlation models on running processes, to identify the possibility of exceptions or undesired behavior. As in the case of the analysis, predictions can be made at the IT level (e.g., predicting whether a given computer or applications will be involved in the execution), or at the business level (e.g., predict whether a service will be delivered in accordance with the stipulated SLA).

Monitoring: BC can monitor and analyze running process instances, and inform the user of unusual or undesired situations. Users can view the health status of the system, processes, services, and resources. In addition they can define critical situations (alerts), so that BC can notify them on the medium of their choice in the event such a critical situation occurs.

Control: based on monitoring and prediction, BC can interact with the BPMS to avoid (or reduce the impact) of foreseen and actual quality degradations, and in particular to help avoid missing SLAs.

Optimization: BC can identify areas of improvement in business process definitions and in the assignment of resources and services to work activities.

Our Contribution

Click to view larger image of this chart.		We have developed Business Cockpit as a prototype implementation that demonstrates how to address the issues mentioned above. Business Cockpit enables users to define, monitor, and measure business-level metrics on top of any data source. Its main features are:

Flexibility: it can provide metrics on top of any data model without requiring ANY change to the code.

Simplicity: everything is performed through point-and-click interfaces. it is suitable for business users.

Power: it enables the definition and analysis of a wide variety of business metrics and service level agreements in a very flexible way.

Performance: the model and system is designed in order to maximize performances both in the metric computation phase and in the analysis phase.

Contact: Ming-Chien.Shan@hp.com