文档介绍：The Schedule as Independent Variable (SAIV) Process for Acquisition of Software-Intensive Systems
Barry Boehm, Winsor Brown, LiGuo Huang
University of Southern California
Center for Software Engineering
Los Angeles, CA 90089-0781
Daniel N. Port
University of Hawaii at Manoa
Honolulu, HI 96822
Abstract
Many system acquisitions do not achieve on-time delivery because of delays in software development. This paper presents a highly essful approach for on-time delivery of software-intensive systems: the Schedule As Independent Variable (SAIV) process. The SAIV process involves prioritization of desired features; scoping a Core Capability of top-priority features easily achievable within the available schedule; architecting for ease of dropping or adding borderline-priority features; monitoring progress with respect to plans; and adding or dropping borderline-priority features to meet the schedule target. The paper summarizes experiences and discusses critical ess factors in applying the SAIV acquisition process across a range from small in-house e-services projects to very large Government systems of systems.
Introduction
For plex systems, software is ing the primary driver of system functionality, cost, and schedule. This is true partly because “software is increasingly the portion of the system that enables the unique behavioural aspects of the system” and because it enables rapid adaptability to change (Maier-Rechtin 2000). It is also true partly because plexity of software makes the duration of its development schedule both longer than the other portions of the system and easier to underestimate.
This latter problem of software schedule achievability has caused many serious difficulties in the acquisition of software-intensive systems. This particularly true for systems whose initial operational capability has been put on contract and is needed by a relatively fixed date. As the date approaches and it es clear that the required functionality will not be