56202

Evolving the Ulysses Mission Control System A Long Term Support Case Study: 

Evolving the Ulysses Mission Control System A Long Term Support Case Study Khee Chan SciSys Ltd, Pasadena, USA; khee.chan@scisys.co.uk David Hall SciSys Ltd, Pasadena, USA; david.hall@scisys.co.uk Christopher Leech SciSys Ltd, Bristol, UK; christopher.leech@scisys.co.uk SpaceOps 2006, Paper AIAA-2006-5693

Mission Milestones: 

Mission Milestones

High Level Data Flow: 

High Level Data Flow

UMCS Through The Ages: 

UMCS Through The Ages

UMCS mark I - Operations Experience: 

UMCS mark I - Operations Experience In service: 1989 – 1997 Operational issues: 100% CPU usage partially alleviated by adding second CPU & recoding SCMS graphics middleware for target display devices Barely sufficient on-line disk storage Major system problems: 1995/6: memory problems in real-time & stand-by systems 1996: CPU failure in real-time system

UMCS mark II - Operations Experience: 

UMCS mark II - Operations Experience In service: 1996 – 2006 New & revised practices: Stockpile hardware spares and source from second-hand market Pay-per-call hardware maintenance instead of flat-fee subscription Operational issues: Worn backup tapes and tape drives Major system problems: Failed tape drive hung SCSI bus on stand-by system Bug in SCMS inter-process communication module

UMCS Options for Continuing Operations: 

UMCS Options for Continuing Operations UMCS mark II Continue with old hardware and software UMCS mark III New hardware Similar but updated operating system Same application software “New heart, same soul” “New” UMCS New hardware New operating system New application software

UMCS mark II – Pros & Cons: 

UMCS mark II – Pros & Cons Pros: No software or hardware changes Cheapest option Cons: Decreasing system availability in future years as components age Coinciding with more severe spacecraft operational constraints Increasing maintenance costs in future years Limited system ”headroom” (CPU, on-line storage etc) to implement new requirements VAX spare parts getting scarce & expensive Riskiest option

UMCS mark III – Pros & Cons: 

UMCS mark III – Pros & Cons Pros Newer & more capable hardware Greater potential system availability in the long term Ample “headroom” for new requirements Same but updated operating system Minimal changes in application software layer (SCMS & FDS) No additional manpower required Straightforward system reverification and revalidation No user retraining while mission operations become less forgiving of human errors Moderate cost and moderate risk Cons Potential scarcity of Alpha spare parts in future years but not as severe as VAX parts

“New” UMCS – Pros & Cons: 

“New” UMCS – Pros & Cons Pros All brand-new hardware Low cost and more capable No maintenance issues expected due to age and more “generic” hardware components More modern application software layer New and hitherto unforeseen functions for improving mission operations Cons High manpower costs due to: Adaptation of application software to JPL and Ulysses unique environment Extensive on-site system integration, verification and validation Major paradigm shift for team while mission operations become less forgiving of human errors Extensive user retraining required

Evaluation of UMCS Options: 

Evaluation of UMCS Options

UMCS Options – Cost Ratings: 

UMCS Options – Cost Ratings

UMCS Options – Risk Ratings: 

UMCS Options – Risk Ratings

UMCS Options – Cost X Risk Ratings: 

UMCS Options – Cost X Risk Ratings

Conclusions 1: 

Conclusions 1 High system availability is more important than high system performance. Operating system maturity and stability is a prime factor in providing long term support. Allow for a hardware upgrade path during planning of a long term project. Hardware is cheap Manpower is expensive Software redevelopment and revalidation is even more expensive

Conclusions 2: 

Conclusions 2 For system migrations: Truly understand the requirements and practices of the mission operations team Human factors must be taken into account Risk to on-going successful mission operations must be quantified and included in trade studies Mission control system support personnel should be part of the mission operations team: Become personal stakeholders in the continued success of the mission Minimise the loss of “corporate memory” during long duration mission

Acknowledgements: 

Acknowledgements Work performed by SciSys Ltd under contract to ESA Ulysses mission logo courtesy of ESA Title slide & slide background artwork courtesy of Bill Mogensen (Ulysses spacecraft controller)