• Version
• 221 Downloads
• 120.76 KB File Size
• 1 File Count
• August 14, 2016 Create Date
• August 14, 2016 Last Updated

Appendix M; Software Cost Estimating

Appendix M: Software Cost Estimating

Software represents a substantial portion of the cost for space systems. Estimating the cost, schedule,
and effort associated with a proposed software development project is a challenging task.

Although software estimation is unique, the cost estimating process described in this handbook still
applies. The primary difference between costing software and hardware or systems is that the dominant
cost component is labor, therefore correctly estimating the development effort is key. The estimation
methods will depend on the resources available and the level of understanding of the needs and
objectives (Task 1) and the ground rules and assumptions (Task 4). A comprehensive process for
software estimation is documented in Jet Propulsion Laboratory’s (JPL’s) Software Cost Estimation
Handbook (JPL-D-26303).

Currently, NASA has agency-wide licenses for both PRICE1 and SEER2 estimating suites, which both
include software estimation tools (see Appendix E). These two specific tools trend toward the higher side
of the cost-complexity spectrum, but there are several other models available to estimate software costs.
One commonly used model is the Constructive Cost Model3 (COCOMO), which was developed by the
Center for Software Engineering (CSE) at the University of Southern California, headed by Dr. Barry
Boehm4.

This appendix will be developed in more detail, but the references should provide the analyst with the
basics to help with software cost estimating. Contact CAD at [email protected].

1 http://www.pricesystems.com/en-us/offerings/pricecostmodels.aspx

2 http://galorath.com/products/software/SEER-Software-Cost-Estimationpr

3 http://csse.usc.edu/csse/research/COCOMOI|/cocomo_main.htm|

4 Boehm, et al, Software Cost Estimation with COCOMO ll, Prentice Hall, 2000.

Appendix M
Software Cost Estimating M-1 February 2015