Predictive business process monitoring aims at forecasting potential problems during process execution before they occur so that these problems can be handled proactively. Several predictive monitoring techniques have been proposed in the past. However, so far those prediction techniques have been assessed only independently from each other, making it hard to reliably compare their applicability and accuracy. We empirically analyze and compare three main classes of predictive monitoring techniques, which are based on machine learning, constraint satisfaction, and Quality-of-Service (QoS) aggregation. Based on empirical evidence from an industrial case study in the area of transport and logistics, we assess those techniques with respect to five accuracy indicators. We further determine the dependency of accuracy on the point in time during process execution when a prediction is made in order to determine lead-times for accurate predictions. Our evidence suggests that, given a lead-time of half of the process duration, all predictive monitoring techniques consistently provide an accuracy of at least 70%. Yet, it also becomes evident that the techniques differ in terms of how accurately they may predict violations and nonviolations. To improve the prediction process, we thus exploit the characteristics of the individual techniques and propose their combination. Based on our case study data, evidence indicates that certain combinations of techniques may outperform individual techniques with respect to specific accuracy indicators. Combining constraint satisfaction with QoS aggregation, for instance, improves precision by 14%; combining machine learning with constraint satisfaction shows an improvement in recall by 23%.

/pdf/comparing-and-combining-predictive-business-process-cxcyu8h9r8.pdf

Comparing and Combining Predictive Business Process Monitoring Techniques

Future service technology will provide an unprecedented access to operational data, which opens up novel opportunities for monitoring, controlling and managing service- oriented business processes. Amongst these opportunities, we consider predictive monitoring to be a major lever for increased efficiency, effectiveness and sustainability in future business networks. Predictive monitoring means that critical events, potential deviations and unplanned situations can be anticipated and proactively managed and mitigated along the execution of business processes. This paper demonstrates the potential of predictive monitoring in practice. We focus on transport a logistics as a major industry sector -- accounting for between 10% to 20% of a country's Gross Domestic Product. Based on widely adopted standards and real operational data, we empirically support the relevance of key issues faced in that industry sector, such as late cancellations of transport bookings and delayed deliveries. As a solution, we describe the design of a novel, cloud- and services-based collaboration and integration platform. Based on this platform we develop short-term prediction capabilities allowing to proactively manage and mitigate the identified issues in the transport a logistics industry, thus promising to increase business efficiency and sustainability.

Predictive Monitoring of Heterogeneous Service-Oriented Business Networks: The Transport and Logistics Case

Services are quintessential in the current economical landscape. Enterprises and businesses at large rely on the consumption and providing of services to ensure their operations and to realize their business offers. That is, nowadays businesses all over the world are interconnected with each other by complex service-centric webs called service networks. The ubiquity and pervasiveness of service networks call for models, methods, mechanisms and tools to understand them and harness their potential. This paper investigates the modelling of the service networks with a focus on busi- ness relationships and exchanges of software services among the involved parties. The contribution of this work is threefold. Firstly, we provide an overview of what ser- vice networks modelling can offer in combination with Business Process Management (BPM) and Service Oriented Architecture (SOA) technologies. Secondly, we propose a formalism to model service networks that depicts them as aggregations of participants - e.g. enterprises or individuals - that offer, request, consume and provide services to each other. With the goal of providing a foundation for the alignment between service network- and business process models, we finally map the constructs of our service networks modelling formalism to the ones of the Business Process Modelling Notation (BPMN).

/pdf/service-networks-modelling-an-soa-bpm-standpoint-2cr80zvhqk.pdf

Service Networks Modelling: An SOA & BPM Standpoint

Networks of interdependent organizations cooperate to produce goods or, nowadays, services that are of value to their markets as well as to the participating organizations. Such co-operations can be supported by corresponding business processes which are based on SOA technology. Developing and managing SOA-based business processes in such service networks necessitates a comprehensive architecture which is on the one hand grounded on solid design principles, and on the other hand capturing best-practices and experiences. Such an architecture is currently lacking. This paper outlines a first attempt to develop and validate an architecture for developing, monitoring, measuring and optimizing SOA-enabled business processes in service networks. A case study from the telecommunications industry is analyzed, and different aspects of service networks are addressed.

An Architecture for Managing the Lifecycle of Business Goals for Partners in a Service Network

We define a service delivery system as a set of interacting entities that are involved in the delivery of one or more business services. A service operating system manages the processes and resources within a service delivery system. This paper presents our on-going work on developing a formal model for these concepts, with the goal of clearly and precisely describing the delivery behavior of service systems. The model lays the groundwork for reasoning about the scenarios that occur in service delivery.

A Formal Model of Service Delivery

The economic structure of service systems has steadily increased in complexity in recent years. This is due not only to specialization in direct material production and services offered, but also in the ownership and management of resources, the role of intangible assets such as process knowledge, and the context in which goods and services are consumed. This increase in complexity represents both a challenge and an opportunity in a service-oriented economy. In this paper, we offer a descriptive structure for the analysis of this complexity which combines graph theory and network flows with economic tools. Our analysis is based on publicly observable information and can be used to analyze service systems in terms of the value they deliver, how they deliver it, and how value can be discovered and increased. We show how this analysis can be applied (in the example of a car manufacturer and its service system for suppliers and dealerships) to improve customer satisfaction and provide options and analysis models for outsourcing decision makers.

G. Iacovidis

Papers

Estimating value in service systems: a case study of a repair service system