Romero Oglietti SSC07

Space Operations as a Guide for a Real-World Scheduling Competition: 

Space Operations as a Guide for a Real-World Scheduling Competition Eduardo Romero (eromero@conae.gov.ar) Marcelo Oglietti (marcelo.oglietti@conae.gov.ar) CONAE (Argentine Space Agency) Scheduling a Scheduling Competition Workshop -ICAPS ´07-

Scheduling Competition: 

Scheduling Competition One of the purposes of the scheduling competition should be bringing theory and applications closer. How? Adding features of real-world problems to the benchmark problems of the competition. What features? We tackle this issue by describing three real-world problems of satellite missions with different characteristics not present in classical scheduling problems.

Outline: 

Outline Satellite Mission Ground Segment Ground Station Services Mission Operation Center Services User Ground Segment Services Characteristics of the problems Conclusions

Outline: 

Outline Satellite Mission Ground Segment Ground Station Services Mission Operation Center Services User Ground Segment Services Characteristics of the problems Conclusions

CONAE Satellite Mission Ground Segment: 

CONAE Satellite Mission Ground Segment (0) Tele Commands, Housekeeping Telemetry, Instruments Data. (1) Request and delivery of instruments raw data. (2) Tele Commands, Telemetry HK (RT & Stored), Scheduling, RT Monitor (3) Requests of science satellite data acquisitions. (4) Satellite Product Requests & Delivery. Organization based on functionality and on responsibility scope.

Outline: 

Outline Satellite Mission Ground Segment Ground Station Services Mission Operation Center Services User Ground Segment Services Characteristics of the problems Conclusions

Ground Station Services: 

Ground Station Services Science Data Download Services Telemetry Tracking and Command Services (TT&CS) GSS receive requests of satellite contacts attendance. These are of several types, requiring the use of different sets of hardware and software units, during specific moments. * = multiple instances

Ground Station Services Scheduling Problem: 

Ground Station Services Scheduling Problem Given a set of ground services requested allocate the use of the units of the ground station in time lines (threads of execution). Similar to the project-scheduling but with extra ingredients. Many of the units have to be instantiated with a set of parameters. Constraints between the use of the units depend on the parameters. An example of a macro configuration (high level). All units have different variables that determine their mode of operation.

Outline: 

Outline Satellite Mission Ground Segment Ground Station Services Mission Operation Center Services User Ground Segment Services Characteristics of the problems Conclusions

Mission Operation Center Services (MOCS): 

Mission Operation Center Services (MOCS) Satellites composition. Subsystems: command and data handling, power, attitude control, mass memory. Payloads: radars, cameras, sensors. MOC receives requests for the use of the payloads and subsystems. Satellites are commanded through sets of basic commands, setting each payload and subsystem.

MOCS Scheduling Problem: 

MOCS Scheduling Problem It consists in three linked scheduling problems with their associated constraints.

Payloads and Subsystems: 

Payloads and Subsystems The activities needed for the two parts of a camera (MMRS) acquisition (high level).

Outline: 

Outline Satellite Mission Ground Segment Ground Station Services Mission Operation Center Services User Ground Segment Services Characteristics of the problems Conclusions

User Ground Segment Services: 

User Ground Segment Services Users request satellite products that have to be processed from raw science data. Production systems have multiple installations in several computers. Each product needs several production systems in order to be processed. Similar to classic job-shop scheduling problems, but priorities of the requests are considered.

Outline: 

Outline Satellite Mission Ground Segment Ground Station Services Mission Operation Center Services User Ground Segment Services Characteristics of the problems Conclusions

Characteristics of the Problems: 

Characteristics of the Problems Dynamic re-scheduling When new services are requested. In case of units failure an alternative schedule has to be provided. Has to be done minimizing the changes of previous schedules (this could be used as a quality measure of the re-scheduling process).

Slide17: 

Distributed scheduling The three problems and their sub problems are linked but cannot be centrally solved because of the scope of responsibilities. Robust scheduling Constant refresh with more exact parameters (e.g., TLE orbit data) demands robust schedules that tolerates minor changes preserving their consistency. Priorities and preferences of the requests arbitrarily imposed

Outline: 

Outline Satellite Mission Ground Segment Ground Station Services Mission Operation Center Services User Ground Segment Services Characteristics of the problems Conclusions

Conclusion: 

Conclusion Add these features to the benchmark problems Use these as Benchmark Problems (?) We propose to build a benchmark problem around these rich real-life scheduling problems. It will cost time and effort, but it will help bring theory and applications closer. Data of the problems can be shared. We presented in the paper a modeling approach that has allowed us to capture many features of these domains