Data-Driven Maritime Processes Management Using Executable Models

Reference Petri Nets, dynamic reconfigurability, executable models, model-driven development, formal specifications

Maritime processes as well as the other logistics-based processes are strongly connected with the mechanical devices they operate on, as well as people involved. Typical operational issues like vessel management, berthing process, cargo loading, etc., as well as the whole predictability of the process fulfilment therefore strongly depends on the processes automation means used. Some of the recent approaches utilise the BPMN processes description for more precise and predictable management of processes. Nevertheless these approaches lack a deeper coverage of technical resources used within the process tasks fulfilment. This paper aims to describe the usage of distributed control system architecture based on Reference Petri Nets models for logistics processes management and decision support. The architecture is based on Reference Petri Nets model covering the service layer of process management, which is used by the BPMN models. System model model consists of the set of involved nodes and their detailed logical functionality description and communication channels. According to predefined set of rules the Reference Petri Nets executable model is generated from the system Specification. This model could be used for the logistics process simulation and estimation, as well as translated into the interpretable Petri Nets bytecode that could be dynamically installed on any node of the system for controlling its functionality. When there are some changes within the model later, they are converted to produce new versions of particular bytecode representing the partial functionality of involved nodes. The important features of described approach are formally precise while human readable description of each layer of the system functionality and the dynamic reconfigurability of target system behaviour during its runtime.