Sustainability of Flight Software for Multi-Generational Time Periods

NASA and other organizations are beginning to seriously consider missions that will extend over such long periods of time that the teams that develop, maintain, and operate their assets will have a complete replacement of staff over the life of the mission. What are the impacts on sustaining and maintaining flight software (FSW) over long time periods? FSW development already suffers from staffing shortages because the programming languages taught in schools are not typically used for FSW; thus, new graduates tend to have limited exposure. Furthermore, languages such as C are now sufficiently out of vogue that new graduates may see investing time to become proficient using them as undesirable in the long run for their career. Staffing is only one concern. A related concern is sustainability of the development and test environment. Languages and operating systems are replaced over time. What if the FSW development environment is no longer supported in a few years? This situation is exacerbated by current cybersecurity practices, a fundamental tenant of which is tools no longer supported must not be used by the organization. Another area of concern is the design of the FSW itself. What will be the necessary practices to make the software flexible, reliable, and maintainable for decades? What might be the role of artificial intelligence? Many current missions tend to see the number of updates fall off dramatically after the primary mission ends in a few years to a decade. This approach seems untenable for a primary mission lasting many decades. We explore these various concerns and suggest ways to deal with each. Finally, we will include early, middle, and late career perspectives from our authors.