Software Process: Applying Industrial Strength Methods In Engineering Education

Abstract

Improving productivity and quality in software development is one of the major concerns of the software engineering discipline, as software systems grow to millions, and soon billions, of lines of code. Productivity and defect density levels that are considered very good today will be inadequate to keep up with this future growth. As a result, software development professionals and organizations are striving to improve existing processes and to develop new ones. One example of demonstrated benefits from these efforts is the success of the Personal Software Process (PSP) and the Team Software Process (TSP) developed by the Software Engineering Institute (SEI) at Carnegie Mellon University. This paper discusses the structure of the TSP and PSP processes, industry experiences with their application, and their relationship to other process improvement frameworks, including the current version of the SEI’s Capability Maturity Model (now known as CMMI). With this background, it reflects on the introduction of the PSP and TSP into software engineering curricula, and reports experiences at the Milwaukee School of Engineering, one of the first four ABET-accredited software engineering programs. Introduction With the introduction of baccalaureate software engineering programs around the world, and the accreditation of the first such programs in the United States, software engineering has become accepted as a full-fledged engineering discipline. This has occurred in the context of the critical importance of software to the global economy and society, and also reflects the role that software now plays in virtually all kinds of engineered systems. As with other emerging areas of engineering, software has had its share of difficulties and failures, but a growing body of knowledge provides a basis for current practice and for continuing future improvement. As is the case in other engineering disciplines, software engineers must have appropriate knowledge and skill both in practice (what they do) and process (how they do it). Software engineering practice has many components, including requirements analysis, software architecture, design, implementation, verification and validation, application of formal and mathematical methods, adaptation to specific application domains, and underlying computer science foundations. Software engineering process addresses issues such as planning, estimation, quality management, teamwork, and continuing improvement of methods and techniques. The purpose of this paper is to describe some areas of software engineering process, including approaches that have demonstrated effectiveness in industry, to report on how similar techniques have been successfully implemented in an undergraduate software engineering program, to reflect on some lessons learned, and to assist other software engineering educators in related efforts.

Proceedings of the 2005 American Society for Engineering Education Annual Conference & Exposition Copyright © 2005, American Society for Engineering Education