SmartPM: An Adaptive Process Management System for Executing Processes in Cyber-Physical Domains

Nowadays, the automation of business processes not only spans classical business domains (e.g., banks and governmental agencies), but also new settings such as healthcare, smart manufacturing, domotics and emergency management [2]. Such domains are characterized by the presence of a Cyber-Physical System (CPS) coordinating heterogeneous ICT components with a large variety of architectures, sensors, actuators, computing and communication capabilities, and involving real world entities that perform complex tasks in the "physical" real world to achieve a common goal. In this context, Process Management Systems (PMSs) are used to manage the life cycle of the processes that coordinate the services offered by the CPS to the real world entities, on the basis of the contextual information collected from the specific cyber-physical domain of interest. The physical world, however, is not entirely predictable. CPSs do not necessarily and always operate in a controlled environment, and their processes must be robust to unexpected conditions and adaptable to exceptions and external exogenous events. In this paper, we tackle the above issue by introducing the SmartPM System (http://www.dis.uniroma1.it/smartpm) an adaptive PMS which combines process execution monitoring, unanticipated exception detection (without requiring an explicit definition of exception handlers), and automated resolution strategies on the basis of well-established Artificial Intelligence techniques, including the Situation Calculus and IndiGolog [1], and classical planning [3].