| Abstract: |
Today’s world economic situation has become an opportunity for many organizations to implement new mechanisms and protocols, unused for them before, in order to increase their productivity and quality of products and improve their services, without increasing production costs. One of the most widely used management strategies to achieve these goals is BPM («Business Process Management»). Over the last decade, BPM seems increasingly oriented towards this direction because the implementation of BPM improves the overall knowledge as well as the «Know-how» at any organization.
In the case of the Software Industry (SI), this one is currently applying BPM as a mechanism to control and define how to build and manage software in order to improve competitiveness. However, BPM in the software context is not as simple as it seems to be because of inherent properties of software processes which are constantly evolving, usually incorporate new lifecycles and technologies, or they are strongly influenced by many work teams, among other aspects. These reasons have caused that SI usually focus on defining their processes, but each involved role performs orchestration – i.e., centralized and coordinated management of events during the process execution – and execution processes in an independent and manual way. For this reason, software processes maintenance, evolution, monitoring and measurement become difficult tasks.
This paper is motivated by these problems in SI to execute and orchestrate business processes. However, although these motivations are framed in SI, we have proposed a flexible solution that can be applied to other areas, for instance, to Health environment, and more particularly, to clinical processes management.
Our proposal1 is to solve the cited problems and facilitate of software processes maintenance by taking advantage of the MDE («Model-Driven Engineering») paradigm. We aim to define domain specific languages needed to model and execute processes and we’ve established systematic transformation protocols among those models. In addition, we tend that all these models and techniques used to model, are intuitive and easily understandable cognitively by non-experts users without technical background. This feature allows these users may participate in the final validation of models.
In this context, our proposal theoretically defines a set of metamodels and systematic QVT and MOFM2T transformation rules. This theoretical framework has been implemented in practice as PLM4BS («Process Lifecycle Management for Business-Software») framework, a CASE tool that provides support to software process lifecycle management in real projects.
PLM4BS solves a particular problem happening at SI: the need for effective, systematic and automated mechanisms to execute and orchestrate software processes from their own definition in order to manage the software-product lifecycle. It is interesting to mention that PLM4BS has been successfully validated in different real environments and confirms that MDE is a promising paradigm in process engineering. Our proposal also lays the theoretical foundation of the EMPOWER platform which allows managing process from a MDE point of view. At present, EMPOWER is being developed by Servinform (a Spanish software company) and has a budget of 700.000€ for 2015-2016.
Regarding future work, it is important to define mechanisms for continuous improvement. So, our future work leads towards answering how the software process monitoring can be included in PLM4BS by means of the MDE paradigm. Other ones are related to simulate and testing processes as well as apply PLM4BS in other business contexts.
1 J. A. García-García. 2015. Una propuesta para el uso del paradigma guiado por modelos (MDE) para la definición y ejecución de procesos de negocios. PhD thesis. University of Seville. URI: http://hdl.handle.net/11441/26740.
|