Larry Putnam Sr. - Why good Measurement, Estimating and Control Techniques are Important for Process Improvement Programs

April 7, 2005

Location: University of Minnesota, EE/CS Building, 3-230

Thursday,
April 7, 2005
5:30-8:00 p.m. at The University of Minnesota
5:30 start for networking, 6:15 start for meeting.

Note: Not the standard room

EE/CS Building
Room 3-230
Minneapolis, MN

Room 3-230 is located in the “right wing” of the EE/CS Building.

Very easy to find, consult the map link below.

Directions:
A map is available at http://onestop.umn.edu/Maps/EE/CSci/
Check out the detailed map under the “close up” button.

This Month’s Meeting:

Program Manager: Jesse Freese

Topic: Why good Measurement, Estimating and Control Techniques are Important for Process Improvement Programs
Speaker: Lawrence H. Putnam Sr.

Abstract:
After 30 years or more, software development continues to experience large cost overruns and schedule slippages despite the many innovations in tools and development processes. Often the reason is that realistic estimates are not done at the time that major decisions are made about undertaking the project. After the project is underway, frequently with an inappropriate budget and schedule, slippage and overrun become apparent. Frantic efforts to adjust to reality are instituted but usually fail.

People at QSM have studied this problem for 27 years. We have examined data from more that 7000 completed projects to learn the real behavior pattern of projects. This has lead to development of good estimating techniques to use at the right time in the business cycle related to the software development project. This presentation brings out the key features of software project estimating early on in the process, at key decision points in the development and how to adaptively forecast changes in plan while underway when it becomes apparent that things are not going according to the original plan. Incidental to estimating is the technique of calibrating the estimating model with a little historic data. This produces a Productivity Index. This index is a good baseline to measure process improvement. So each time a project is completed and a new estimate made, a process improvement measurement is obtained. Over time we learn what our rate of improvement is and how our organization compares with others in the industry.

Several case studies will be used to illustrate how estimating and control ideas have been used to manage successful projects and to measure process improvement. The benefits of following these practices are delivering better products on time within budget, making better business decisions and at the same time being able to measure and control the organization’s process improvement objectives.

Lawrence H. Putnam is the founder and CEO of Quantitative Software Management, Inc., a developer of commercial software estimating, benchmarking and control tools known under the trademark SLIM®. He served 26 years on active duty in the U.S. Army and retired as a colonel. He holds a BS from the United States Military Academy, a MS degree in Physics from the Naval Postgraduate School. He has been deeply involved in the quantitative aspects of software management for the past 30 years. He is the co-author of five books on software estimating, control and benchmarking. He is a member of Sigma Xi, ACM, IEEE and IEEE Computer Society. He was presented the Freiman Award for outstanding work in parametric modeling by the International Society of Parametric Analysts.

The following files related to this presentation are available for download:

Slides from Presentation
Type: application/pdf, Size: 774 KB
Description: