Why are Space Stations so Hard?

1
Why are Space Stations so Hard?
8th Military and Aerospace Programmable
Logic Devices (MAPLD) International Conference 8
Sept. 2005
Bill Dwyer, NASA-JSC ISS Command and Data
Handling System Hardware System
Manager Ex-Space Station Freedom Data Management
System Integration Manager Ex-Space Station
Freedom SW Functional Area Manager Ex-Space
Station Freedom Mode Team Chairman (Level-II)
2
Development Environment
1

• Extended development phase some articles are
  on-orbit while others that will
• interface to it are still in an early phase
  such as PDR or in some cases not started yet
• Makes interface definition very challenging
• Requiring extensive on-ground integration
  testing is a large complex test rig
• Widely distributed development sites
  International world site distribution
• Directly contradicting co-location axiom for
  project development
• Again makes interface definition and testing
  very difficult
• After the first two or three element launches,
  a system can be in multiple lifecycle
• stages at one time
• Sustaining, on-orbit troubleshoot and root
  cause work and upgrade development
• This multiple-lifecycle-at-one-time state does
  not allow the post-development
• resource phase out that is the standard model
• International export-import regulations serve
  to extend time required for testing and
• some aspects of development

3
On-Orbit Operations
2

• Coordinating flight control across
  International boundaries
• To date, ISS has had a great deal of success in
  International operations
• Substantial planning and agreements required
• On-orbit troubleshooting and repair of failures
  has not been done to the scale it is
• being accomplished on ISS
• Earlier Mir experience very valuable
• Command and Control computational systems
  failures during 6A flight timeframe
• Turns the crew into multi-skilled mechanics
• On-orbit assembly and checkout of major
  components, including multi-national assembly
• Attaching new element such as truss element S1
  and P1 to existing configuration
• Upgrading to a new integrated command and data
  handling system with the
• addition of multiple new computational
  systems
• On-orbit change-out of hardware, especially
  external hardware
• Example CMG failure and change-out on recent
  flight LF-1 (STS-114)
• New, unforeseen requirements/needs that must be
  implemented on-orbit with minimal

4
On-Orbit Operations
3

• Installation of major software systems on-orbit
  with no operational fidelity sacrifice
• A design for requirement that is ops
  intensive

• Fixing whoopses requirements that were not
  anticipated but need arises during ops
• For example the 6B box bowing problem for CT
  and CDH HW
• Relocation of major systems due to operations
  or safety needs
• For example the Acvanced ECLSS in the U.S. Lab
• Emergent behavior of large complex systems
  sometimes things behave in an
• unanticipated manner
• For example restarting large networks after
  flight 6A command and control
• computational system failures
• Difficulty in determining operational life of
  electronic hardware makes logistical
• planning very difficult
• Some good the ISS computational systems,
  multiplexer-demultiplexers (MDMs)
• have proven to be
  extremely robust, some operating nearly 8 years
• with no problems
• Some not-so-good laptop inventory needs
  complete replacement

5
Conclusion
4

• The devil is in the details
• The devil is in the integration
• The devil is in the magnitude of the
  undertaking
• The devil is resource intensive and costs a lot
• Building a large multi-element, multi-national
  space station is
• indeed hard and expensive. If we underestimate
  the magnitude
• of the task, it can, will and has given us more
  than one negative
• surprise.