Raymond J. Madachy

Cost models for future software life cycle processes: COCOMO 2.0

Current software cost estimation models, such as the 1981 Constructive Cost Model (COCOMO) for software cost estimation and its 1987 Ada COCOMO update, have been experiencing increasing difficulties in estimating the costs of software developed to new life cycle processes and capabilities. These include non-sequential and rapid-development process models; reuse-driven approaches involving commercial off-the-shelf (COTS) packages, re-engineering, applications composition, and applications generation capabilities; object-oriented approaches supported by distributed middleware; and software process maturity initiatives. This paper summarizes research in deriving a baseline COCOMO 2.0 model tailored to these new forms of software development, including rationale for the model decisions. The major new modeling capabilities of COCOMO 2.0 are a tailorable family of software sizing models, involving Object Points, Function Points, and Source Lines of Code; nonlinear models for software reuse and re-engineering; an exponentdriver approach for modeling relative software diseconomies of scale; and several additions, deletions and updates to previous COCOMO effort-multiplier cost drivers. This model is serving as a framework for an extensive current data collection and analysis effort to further refine and calibrate the models estimation capabilities.

Using the WinWin spiral model: a case study

Fifteen teams used the WinWin spiral model to prototype, plan, specify, and build multimedia applications for USCs Integrated Library System. The authors report lessons learned from this case study and how they extended the models utility and cost-effectiveness in a second round of projects.

A software product line life cycle cost estimation model

Most software product line cost estimation models are calibrated only to local product line data rather than to a broad range of product lines. They also underestimate the return on investment for product lines by focusing only on development vs. life-cycle savings, and by applying writing-for-reuse surcharges to the entire product rather that to the portions of the product being reused. This paper offers some insights based on the exploratory development and collaborative refinement of a software product line life cycle economics model, the Constructive Product Line Investment Model (COPLIMO) that addresses these shortfalls. COPLIMO consists of two components: a product line development cost model and an annualized post-development life cycle extension. It focuses on modeling the portions of the software that involve product-specific newly-built software, fully reused black-box product line components, and product line components that are reused with adaptation. This model is an extension built upon USC-CSEs well-calibrated, multi-parameter Constructive Cost Model (COCOMO) II, tailored down to cover the essentials of strategic software product line decision issues and available supporting data from industries.

System dynamics modeling of an inspection-based process

A dynamic simulation model of an inspection-based software lifecycle process has been developed to support quantitative process evaluation. The model serves to examine the effects of inspection practices on cost, scheduling and quality throughout the lifecycle. It uses system dynamics to model the interrelated flows of tasks, errors and personnel throughout different development phases and is calibrated to industrial data. If extends previous software project dynamics research by examining an inspection-based process with an original model, integrating it with a knowledge-based method for risk assessment and cost estimation, and using an alternative modeling platform. While specific enough to investigate inspection practices, it is sufficiently general to incorporate changes for other phenomena. It demonstrates the effects of performing inspections or not, the effectiveness of varied inspection policies, and the effects of other managerial policies such as manpower allocation. The results of testing indicate a valid model that can be used for process evaluation and project planning, and can serve as a framework for incorporating other dynamic process factors.

Heuristic risk assessment using cost factors

The author describes an expert system tool that can be used to analyze patterns of software cost driver ratings submitted for a Cocomo cost estimate. These results help users determine and rank associated sources of software project risk.

The ROI of software dependability: The iDAVE model

In most organizations, proposed investments in software dependability compete for limited resources with proposed investments in software and system functionality, response time, adaptability, speed of development, ease of use, and other system capabilities. The lack of good return-on-investment models for software dependability makes determining the overall business case for dependability investments difficult. So, with a weak business case, investments in software dependability and the resulting system dependability are frequently inadequate. Dependability models will need to support stakeholders in determining their desired levels for each dependability attribute and estimating the cost, value, and ROI for achieving those. At the University of Southern California, researchers have developed software cost- and quality-estimation models and value-based software engineering processes, methods, and tools. We used these models and the value-based approach to develop an Information Dependability Attribute Value Estimation model (iDAVE) for reasoning about software dependabilitys ROI.

A stakeholder win–win approach to software engineering education

We have been applying the stakeholder win–win approach to software engineering education. The key stakeholders we are trying to simultaneously satisfy are the students; the industry recipients of our graduates; the software engineering community as parties interested in improved practices; and ourselves as instructors and teaching assistants. In order to satisfy the objectives or win conditions of these stakeholders, we have formed a strategic alliance with the USC Libraries to have software engineering student teams work with Library clients to define, develop, and transition USC digital library applications into operational use. This adds another set of key stakeholders: the Library clients of our class projects. This paper summarizes our experience in developing, conducting, and iterating the course. It concludes by evaluating the degree to which we have been able to meet the stakeholder-determined course objectives.

Developing multimedia applications with the WinWin spiral model

Fifteen teams recently used the WinWin Spiral Model to perform the system engineering and architecting of a set of multimedia applications for the USC Library Information Systems. Six of the applications were then developed into an Initial Operational Capability. The teams consisted of USC graduate students in computer science. The applications involved extensions of USCs UNIX-based, text-oriented, client-server Library Information System to provide access to various multimedia archives (films, videos, photos, maps, manuscripts, etc.).

Active learning and effort estimation: Finding the essential content of software effort estimation data

Background: Do we always need complex methods for software effort estimation (SEE)? Aim: To characterize the essential content of SEE data, i.e., the least number of features and instances required to capture the information within SEE data. If the essential content is very small, then 1) the contained information must be very brief and 2) the value added of complex learning schemes must be minimal. Method: Our QUICK method computes the euclidean distance between rows (instances) and columns (features) of SEE data, then prunes synonyms (similar features) and outliers (distant instances), then assesses the reduced data by comparing predictions from 1) a simple learner using the reduced data and 2) a state-of-the-art learner (CART) using all data. Performance is measured using hold-out experiments and expressed in terms of mean and median MRE, MAR, PRED(25), MBRE, MIBRE, or MMER. Results: For 18 datasets, QUICK pruned 69 to 96 percent of the training data (median = 89 percent). K = 1 nearest neighbor predictions (in the reduced data) performed as well as CARTs predictions (using all data). Conclusion: The essential content of some SEE datasets is very small. Complex estimation methods may be overelaborate for such datasets and can be simplified. We offer QUICK as an example of such a simpler SEE method.

Case studies in software process modeling with system dynamics

Littons Guidance and Control Systems (GCS) Division has been using system dynamics to create mostly small-scale models for investigating managerial process issues and supporting personnel training. At the project level, these include models for planning specific projects, studying Brookss Law and hiring issues, an interactive earned value model, requirements volatility and a detailed peer review model. The perspective of some of the models has been at a multi-project or departmental level including domain learning, product-line reuse processes and resource contention among projects. Insights provided by the models have supported decision-making at different levels and helped galvanize process improvement efforts. The training applications have added spark in classes and improved overall learning. The models encapsulate collective knowledge of modeling participants, and support organizational learning. By examining the models and simulated behavior, managers share a process vision and can discuss issues against the common models. The models have helped managers understand the key factors in complex scenarios. Knowledge of the interrelated technical and social factors coupled with simulation tools has enabled GCS to improve their planning and management processes. Modeling is also used to support training of software managers and leads. Topics including earned value techniques, productivity estimation, requirements volatility effects and extrapolation of project tracking indicators have been presented with simulation models. Some of these are ‘flight training’ scenarios that the students interact with to practice project control. Though we are in the early stages of many of these efforts, they will be followed through and reported on at a later time. This paper describes our introductory experiences and plans for the future. Copyright