~~

SCHOOL OF MECHANICAL ENGINEERING SEMINAR
Monday, December 2, 2019 at 14:00
Wolfson Building of Mechanical Engineering, Room 206

PROVE - A framework for development process design

Avi Shaked
PhD Yoram Reich

The development of systems is becoming increasingly challenging with respect to the complexity of the system as well as to the large scale, collaborative and integrative nature of the development effort. The design of system development efforts is, therefore, considered an essential contributor to their success. Unfortunately, development process design remains an undisciplined domain, lacking applicable, research-proven tools as a design infrastructure. Accordingly, the resulting design artifacts – development process plans – are typically flawed, risking the success of development efforts.
In this research, we propose and validate a fit-for-purpose, model based design framework to support development process design, in a way which corresponds with its ontology and contributes to its rigor. First, we derive relevant domain ontology for development process design, revealing desirable characteristics of such design. We then combine these with pertinent aspects of process modeling, and discuss the inadequacy of existing modeling approaches for the domain. Then, we describe a new modeling framework we designed particularly for development process design. This framework – PROVE – comprises both domain specific modeling methodology and domain specific notation, and embodies adaptations of modern modeling concepts and usability features. Next, we report multiple, diverse case studies in which the new framework was employed. These case studies demonstrate PROVE as a valid, applicative modeling framework, which contributes to various aspects of development process design, including its communicability, consistency and reuse. We also use PROVE concepts to extend a general purpose modeling framework (OPM), increasing both the practical usability of the general purpose framework for process design and the rigor of the resulting process models.  After establishing PROVE as beneficial, we conclude by highlighting further potential in its integration with other domain specific perspective as well as by offering possible generalizations of our work with respect to the design and use of modeling frameworks.