Efficient Planning Systems Development with EKLOPS

1
Efficient Planning Systems Development with EKLOPS

• Marc Niézette, Ian Shaw, Erik Noréus, Robin
  Steel, Jörg Noll, Holger Dreihahn, and Helmut
  Schönhardt

2
Presentation Outline

• Why EKLOPS?
• Technical Approach
• EKLOPS History
• Benefits of EKLOPS
• Conclusion

3
Mission Planning System Characteristics

• Frequent change requests
• Highly dependent on numerous changing interfaces
  (Flight Dynamics, Data Relay Satellite MCF, etc.)
• Evolves with the mission
• Specificity of planning algorithms
• Problem modelling sensitive to requirement
  changes
• Short development time frame
• Impact on system validation before deployment
• A Flexible approach is required

4
EKLOPS Components

5
Technical Approach

• Libraries of configurable C components
• Mission Planning domain representation (plan,
  activity, resource, etc.)
• Rule-based engine
• XML-based system configuration
• I/O filters (ingestion of input data, scheduling,
  reporting)
• Human Computer Interface components (Timeline
  display, Resource Profile displays, Etc.)
• Plug-in interfaces for integrating planning
  modules in the application logic

6
Planning Modules in EKLOPS
Requests Implementation
Ref. Orb. TL Generation



Requests Selection

Links Selection Rule

Switches Selection Rule

Visibility

Requests

Links

Switches

Time

7
Mission Planning System Architecture
8
Planning Timeline
9
Planning Timeline (Detail)
10
EKLOPS Evolution
ENVISATFOSMPS
Kernel
EKLOPS 1.0
EKLOPS 2.x
EKLOPS 3.x
APS
MEX MPS
EPS
LMPS (P)
GMPS (P)
VEX MPS
EMS PS (P)
11
EKLOPS 1.0 Envisat Mission

• Planning/Checking of observation requests
• Data Management System planning (Solid State
  Recorders, links, acquisitions, etc.)
• Spacecraft and station schedule generation
• Process architecture
• Static and dynamic data representation
• Basic HCI components
• Plug-in framework
• Available on Solaris

12
EKLOPS 2.x MEX/VEX

• Orbiter operations constraint checking
• Complex modelling of resources (thermal, power,
  memory)
• Spacecraft and station schedule generation
• Full XML configurability
• External XML formats for all data
• Interface with EGOS MCS infrastructure
  (SCOS-2000)
• Improved HCI components
• Available on Solaris

13
EKLOPS 3.x APS/EPS

• Science plan optimization
• Network service allocation
• Flexible plan representation
• Requests specified as high-level planning goals
• Logical query language
• Integration of search and optimization algorithms
• Available under LINUX

14
EKLOPS 3.0 Open Architecture
GUI JAVA Components
Task Proxy
Framework Task
Database
Application Control
Output Manager
Planning Plug-in Interface
Input Plug-in Interface
I1
I2
In
P1
P2
Pm
15
Benefits of EKLOPS

• ESA IPR (operational software available at no
  license fee to ESA members)
• Development benefits from re-use of configurable
  components
• Minimum use of COTS
• Validation effort reduced
• Low operation costs (automation)
• Maintenance cost reduced and shared with other
  customer/projects
• Training cost reduced if several missions are
  based on the kernel
• Kernel validated and used in several operational
  missions
• Commonality with ESOC mission operations system
• Easily tailorable and expandable

16
Conclusion

• VEGA has successfully implemented an approach to
  the development of planning systems based on the
  re-use of a kernel library, the Enhanced Kernel
  Library for Operational Planning Systems (EKLOPS)
• EKLOPS has been used as the basis for the
  development of several operational planning
  systems now in operation
• The library evolves and is improved with each new
  development
• EKLOPS provides a cost-effective low-risk
  approach to the development of operational
  planning systems

17
www.vega-group.com
Independent Programme and System
Assurance Technical Excellence . Pragmatic
Solutions . Proven Delivery