Software Engineering for Satellites.pdf

Software Engineering
for Satellites
Kathryn Anne Weiss
Software Engineering Research Laboratory
Department of Aeronautics and Astronautics
Massachusetts Institute of Technology
October 22, 2003
June 17, 2004 © Massachusetts Institute of Technology, 2002 1
Topics of Discussion
‹ Background ‹ Why is Software
Engineering Hard for
 Why is Software
Engineering Hard? Spacecraft?
 Lifecycle ‹ SERL Approach
• Cost ‹ Component-Based
• Requirements Specification Systems Engineering
• Approaches to Design
 SPHERES
• Implementation
• Testing ‹ Conclusions
• Maintenance
June 17, 2004 © Massachusetts Institute of Technology, 2002 2
Background
Ariane 5
Courtesy of Arianespace / ESA / CSG.
Used with permission.
Mars Climate
Orbiter
SOlar Heliospheric Observatory
June 17, 2004 © Massachusetts Institute of Technology, 2002 3
Background
‹ Why is Software Engineering Hard?
 “Curse of flexibility”
•‘‘And they looked upon the software and saw that it was good.
But they just had to add one other feature ...’’
• No physical constraints
 Intangibility
 Lack of historical usage information
 plexity
• plex plete analysis
• anized for statistics
 Large discrete state spaces
June 17, 2004 © Massachusetts Institute of Technology, 2002 4
Background
‹ Software Lifecycle
Feasibility
Study Requirements
V & V Design
V & V Implement
V & V ation
Testing
V & V
V & V Maintenance
V & V
June 17, 2004 © Massachusetts Institute of Technology, 2002 5
Background
‹ Software Cost Coding
Requirements
Testing
Maintenance
June 17, 2004 © Massachusetts Institute of Technology, 2002 6
Requirements Specification
‹ Most critical portion of the software lifecycle
‹ Majority of errors in software can be traced back to
flaws in the requirements
‹ Many methods and types of requirements including:
Informal Formal
•English •Zed
•UML •State Machines
•Intent Specifications
June