FOUR DECADES OF EVOLUTION TOWARD FLEXIWARE (From the perspective of an Apollo, Shuttle, ISS and Shuttle Upgrades Veteran) Definitions

1
FOUR DECADES OF EVOLUTION TOWARD FLEXIWARE(From
the perspective of an Apollo, Shuttle, ISS and
Shuttle Upgrades Veteran)Definitions

• Flexiware My term for an idealized amorphous
  melding of code and electronic devices which
  enable a Flight Computer System to readily
  incorporate design changes necessary to meet
  changes in functional and performance
  requirements.
• Flight Computer System The integrated set of HW
  and SW which provides real time data I/O,
  processing and error protection in support of
  vehicle systems. Note the I/O includes User
  Interfaces, onboard and/or elsewhere.

2
FIRST GENERATION DIGITAL FLIGHT COMPUTERS(1960s)

• The Hardware and Software used in the first
  generation of digital flight computers (1960s)
  were very accurate descriptions of how a
  computer could accommodate changing requirements.
• Hardware components were stacked into sealed,
  encapsulated modules like cordwood
• By contrast, Software was hand-crafted by
  skilled artisans who practically memorize their
  code and could readily tweak it upon request.

3
SECOND GENERATION DIGITAL FLIGHT COMPUTER
SYSTEMS(1970s)

• With the inception of ICs and PROMS, Hardware
  became slightly less difficult to change
• Instruction Set tweaks by PROM R/R, a.k.a.
  Firmware change.
• Minor HW logic changes by delicate placement of
  softwires on multilayer circuit cards
• Conversely, Software became harder to change
• Increases in size and complexity (to meet
  increased scope of applications) moved it beyond
  the domain of (most) artisans
• Memory limitations frequently prevented ideal
  structure and modular design
• Increasing scope and complexity of vehicles, and
  mission requirements, increased risk of
  requirements errors and/or requirements creep.
• Management of HW and/or SW changes in concert
  with ongoing system development/verification
  became a critical element of Project success.
• This above situation prompted 2 signs to appear
  on my Bosss office wall
• Better is the Enemy of Good
• The more innocuous the design change, the more
  far-reaching the consequences

4
THIRD GENERATION FLIGHT COMPUTER SYSTEMS1980s

• Increasing density of ICs and early Gate Array
  devices provided limited additional means for
  hardware design modifications
• Firmware designation extended to include
  Programmable Gate Array Devices
• Space Radiation Effects became a significant
  consideration in selection of EEE Parts
• Application of Software Engineering principles,
  in conjunction with increased memory capacity,
  did allow SW designs which accommodated
  incremental change traffic.
• However, while technology advances in HW and SW
  offer the potential for increased capabilities,
  our ability to generate correct, and stable,
  requirements did not increase commensurately. The
  quotes on my bosss wall may have faded, but they
  were still proven correct more than once.

5
FOURTH GENERATION FLIGHT COMPUTER SYSTEMS((1990s
to present)

• Hardware device technology (FPGA, etc) has
  evolved such that HW (re)design flexibility is an
  essential element in early prototypingto the
  point of offering alternative implementations
  once allocated to SW.
• Mitigation of space radiation effects has become
  (arguably) the predominant influence on EEE parts
  selection and overall FCS architecture.
• Conversely, utilization of COTS Software for OS,
  and other, functions effectively limits SW
  (re)design flexibility in areas that were once
  the province of SW artisans.
• In 40 years, the classical distinction between
  Hardware and Software has become increasingly
  blurred. The net result is increased design
  flexibility that should be exploited for early
  prototyping. Todays increasingly complex
  systems demand thorough and detailed prototyping
  to ensure (1) correct and complete requirements
  and (2) a HW/SW implementation capable of meeting
  all those requirements.