A metamodel to integrate business processes time perspective in BPMN 2.0

Abstract

(ISO/IEC 19510:2013) and selected technology to model processes. The temporal dimension underlies in any kind of process however, technicians need to shape this perspective that must also coexist with task control flow aspects, as well as resource and case perspectives. BPMN poorly gathers temporary rules. This is why there are contributions that extend the standard to cover such dimension. BPMN is mainly an imperative language. There are research contributions showing time constraints in BPMN, such as (i) BPMN patterns to express each rule with a combination of artifacts, thus these approaches increase the use of imperative BPMN style, and (ii) new decorators to capture time rules semantics giving clearer and simpler comprehensible specifications. Nevertheless, these extensions cannot yet be found in the present standard. Objective: To define a time rule taxonomy easily found in most business processes and look for an approach that applies each rule with current BPMN 2.0 standard in a declarative way. Method: A model-driven approach is used to propose a BPMN metamodel extension to address timeperspective. Results: We look at a declarative approach where new time specifications may overlie the main control flow of a BPMN process. This proposal is totally supported with current BPMN standard, giving a BPMN metamodel extension with OCL constraints. We also use AQUA-WS as a software project case study which is planned and managed with MS Project. We illustrate business process extraction from project plans. Conclusion: This paper suggests to handle business temporal rules with current BPMN standard, along with other business perspectives like resources and cases. This approach can be applied to reverse engineering processes from legacy databases.

A rewriting logic approach to resource allocation analysis in business process models

“…To simplify the exposition of the running example (Figure 2), the delays in all flows are set to 0 and the specification of the task duration has been placed at the bottom-left corner of the process description. For instance, the duration of the Sign in task follows a normal distribution with mean 1 and variance 0.5, and the Search products task follows a uniform distribution in the interval [3,30]. Exclusive split gateways are modeled using probabilities associated to outgoing flows.…”

“…By using the gradient descent algorithm on the running example, starting from the assignment (employee → 4, drone → 4), and using the optimization function and values in Section 4.3, the optimal solution can be found by following the sequence (4, 4) −→ (3, 5) −→ (2, 6) −→ (2, 7) −→ (3,8). The values for which the optimization function is computed by our search function and the corresponding evaluations is depicted in Table 3.…”

Stochastic analysis of BPMN with time in rewriting logic

“…Several works propose extensions of BPMN with time constructs, see, e.g. [7,24]. In [24], the authors present Time-BPMN, an extension of BPMN to represent various constraints and dependencies that may arise when modelling business processes.…”

“…The temporal constraints are used to control the beginning/end of an activity, whereas temporal dependencies involve two activities and indicate the relationship between their respective beginning/end. In [7], a metamodel-based approach to integrate temporal constraints and dependencies is introduced. The time aspects are specified using rules and OCL constraints capture the semantics of these rules.…”

“…The time aspects are specified using rules and OCL constraints capture the semantics of these rules. [24] introduces time as a first class citizen to the BPMN standard whereas [7] proposes an MDE-based approach that enrich the existing BPMN control-flow graph model. In comparison, the goal in the present paper was not to extend the BPMN notation to take expressive modeling of time constraints into account, but to rely on usual time representation (duration of task and flow) and concentrate on the analysis of key timing properties in these models.…”