Article Preview
TopIntroduction
In the last few years, with the wide adoption of PAIS (Process Aware Information Systems) such as ERP (Enterprise Resource Planning), companies are left with directories containing several variants of the same business processes. These variants differ according to their application context. For instance, in the e-healthcare domain, 90 variants of “medical examination process” could be distinguished in a hospital (Ayora et al., 2012). Consequently, in order to choose or combine variants, the designer has to compare and adapt them manually, which could be a complex and an error prone operation. In this context, many research studies have focused on managing the variability of business processes by developing configurable processes (La Rosa, 2009; Hallerbach et al., 2010; Reinhartz-Berger et al., 2009; Kumar et al., 2012; Ayora et al., 2013; Nguyen et al., 2014; La Rosa et al., 2017). The variability management is about three main phases: modeling, resolution and evolution. In terms of modeling, many approaches for business process variability representation have been proposed (Schnieders et al., 2007; La Rosa, 2009; Hallerbach et al., 2010). Regarding the resolution, there are several solutions enabling the generation of variants (Gottschalk et al., 2009; Hallerbach et al., 2010; Arriagada-Benítez et al., 2017). Whereas the evolution still presents many limitations, the main contribution in this domain is the work of (Ayora et al., 2016) which is an extension of the change patterns of (Weber et al., 2008) to support changes of process families. However, this proposal is limited to the processes represented in structural blocks and does not assist the designer during the evolution. In this context, we have proposed a pattern-based approach offering model and process solutions (Sbai et al., 2015).
Along with the improvement of business process reuse by the introduction of the variability management, the proposed approaches lack of business-IT alignment support. Our study shows that these approaches do not allow the generation of configurable services emanating from configurable processes, in this context a new concept related to business processes has been introduced, which is the “service-based process model” (Sbai et al., 2014). This has led us to study the alignment between the configurable processes and the enterprise applications, in particular services, with the aim of building PAIS that support service orientation. The emergence of variability management in business processes and services conduct the PAIS today to adopt the configurable processes at the business layer and the configurable services at the IT layer. In this perspective, we proposed an MDA (Model Driven Architecture) approach for the configurable services generation (Wang et al., 2010).