Marc I. Kellner

Software Process Modeling: Principles Of Entity Process Models

A defined software process is needed to provide organizations with a consistent framework for performing their work and im- proving the way they do it. An overall framework for modeling simplifies the task of producing process models, permits them to be tailored to individual needs, and facilitates process evolution. This paper outlines the principles of entity process, models and suggests ways in which they can help to address some of the problems with more conventional approaches to modeling software processes.

Software Process Modeling Support for Management Planning and Control

This paper demonstrates the application of a specific software process modeling approach to certain key managerial activities: ex ante planning, monitoring and recording progress, and dynamic replanning. A realistic example process, with assumed task durations and outcomes, forms the foundation for the illustrations. Automated, quantitative simulations are used to derive schedules, required work effort, and required staffing profiles. Cases of both point estimates (deterministic modeling) and uncertain estimates (stochastic modeling) are discussed, and resource constraints are also considered. This modeling approach is shown to offer distinct advantages over traditional project management approaches such as the critical path method and PERT.

Software process modeling: a case study

Experiences in applying a specific modeling approach and technology to a portion of a software support process used by the US Air Force are related. The modeling approach is discussed in the context of examples drawn from the model developed. A view of the primary objectives of software process modeling, which formed the basis of the approach used, is set forth. The usefulness of the model is evaluated, and general lessons are drawn from the modeling effort.<<ETX>>

Empirical analysis in software process simulation modeling

Abstract Software process simulation modeling is increasingly being used to address a variety of issues from the strategic management of software development, to supporting process improvements, to software project management training. The scope of software process simulation applications ranges from narrow focused portions of the life cycle to longer-term product evolutionary models with broad organizational impacts. This paper discusses some of the important empirical issues that arise in software process simulation modeling. We first address issues concerning real-world data used to (1) establish input parameters to a software process simulation model, and (2) establish actual organizational results against which the model’s results (i.e., outputs) will be compared. On the input side, the challenges include small sample sizes, considerable variability and outliers, lack of desired data, loosely defined metrics, and so forth. On the output side, the paper addresses (1) verification and validation of the model, and (2) quantitative approaches to evaluating model outputs in support of managerial decision making including financial performance using Net Present Value (NPV), multi-criteria utility functions, and Data Envelopment Analysis (DEA). The paper focuses on the stochastic modeling using Monte Carlo simulation. The paper is grounded in the authors’ practical application experiences, and major points are illuminated by examples drawn from that field work.

Experience with enactable software process models

Models of software life-cycle processes provide a means of reasoning about organizational processes used to develop and maintain software. Clearly, these processes play a substantial role in determining the quality, responsiveness, cost, and schedule of a software system. As a result, improvements to software engineering processes should lead to significant improvement in these characteristics. I see software process modeling as an emerging technique, that has enormous potential for contributing to the improvement of software processes and their corresponding products. This paper1 discusses experience in developing and using enactable software process models at the Software Engineering Institute (SEI). Some thoughts on future directions for these efforts are also presented.

Non-technological issues in software engineering

A panel discussion is reported. Three nontechnological problems were selected which each serve to cluster several lower-level elements. First, the software engineering profession has not produced a cadre of capable/competent managers. Second, software development is largely practiced as an individual creative activity, rather than a team effort. Third, the software engineering community has not taken positive action to reduce the performance (e.g., productivity and quality) differences among individuals (or across teams). Much of the potential impact of technological advances may be blocked by these factors. The panelists offer a wealth of experience, insight, and international perspectives on these nontechnological issues. The panelists address both the problems and steps for their resolution.<<ETX>>

Software process modeling example

This paper’ is intended to convey the flavor of our approach to software process modeling through the presentation of an example model fragment. Unfortunately, in the very limited space available for this paper, it is only possible to illustrate some of the key aspects of our modeling approach. The interested reader is referred to [4,6,7] for more comprehensive discussions and examples of our models.

The effects of software processes on meeting targets and quality

Firms developing software face increasing pressures to improve product quality, productivity, time to market, and customer satisfaction. As projects increase in size and complexity, and competition grows, firms are reengineering their software processes. They are adopting more intensive procedures for requirements management, project planning, defect tracking, configuration management, design and code inspections, and so forth. To explore the potential effectiveness of these efforts, we conducted a survey of senior practitioners at the 1993 Software Engineering Process Group National Meeting. The survey asked participants about the processes followed on, and the outcome of a specific software project. Certain practices, notably project planning and cross-functional teams, were consistently associated with favorable outcomes. Based on the survey results, other practices may have little impact on project outcomes.<<ETX>>

Connecting reusable software process elements and components

The paper focuses on one particular topic within the realm of software process reuse: namely, the issue of connections among reusable software process elements and components. Such connections can be problematic even in cases where human ingenuity and understanding is counted on entirely to smooth out the connection details while performing the process. However, the difficulties become more severe in cases involving process technology, such as formal representation of processes in a process modeling language, automated analysis and simulation of processes, and automated execution support for processes (aka process-centered environments). The paper addresses the connection issues that arise across all of these cases, particularly including the challenges that process reuse pose to process technology.

A Reference Model for Process Technology

The reference model described in this paper is a conceptual framework with which we hope to characterize and categorize the field of process technology. We have chosen to partition and decompose the domain in question into a set of defined, related sub-fields. The reference model allows discussion of both the component concepts and the many relationships that exist between those concepts. From a practical perspective, the model also potentially provides a structure through which process technology information and assets can be rapidly retrieved. Currently this model is a work-in-progress. It is hoped that feedback from the process-technology community will help both with its organizational structure and its content.