Optimizing for Responsive Space Design AIAARS20045005 Terrance Yee

1
Optimizing for Responsive Space
DesignAIAA-RS-2004-5005Terrance Yee
2

Introduction

• MSI believes that small experimental satellites
  offer an opportunity to greatly shorten the
  development timeline compared to traditional
  spacecraft
• Rapid design requires significant buy-in from the
  customer including coordination on risk approach,
  required documentation, and most importantly
  ground rules for generating requirements

Streamlined Requirements, Design Documentation
Roadrunner
Terra
3

Pre-Design Phase

• Step 0 Mission Purpose, Team Selection,
  Contracting
• Can be the most incompressible part of a
  responsive mission
• Limited optimization due to regulations, fairness
  requirements
• Alternate method
• Step 0.5 Define Core Purpose, Core Team and
  start with funding just for those
• Gets basic system design started immediately
• Allows long lead hardware to be purchased
• Less cost efficient overall due to necessity of
  adjusting to late additions to the mission
• Constrains later additions to work within
  framework of existing capability

4

Requirements

• Minimizing schedule means setting an early
  baseline design that maximizes use of existing
  elements and short-lead time components
• New payloads participate on a capabilities
  driven basis they are only manifested if they
  can be accommodated within the capabilities of
  the existing design
• Working within the framework of an existing
  design means that performance is often not
  specified it is a function of the
  hardware/software available, not an independent
  variable to be chosen

Max Return w/ Available System
Demonstration Objectives
Requirements
Flexible CONOPS
5

Requirements Minimization

• Dont Duplicate ICDs
• Focus on primary functions, not performance
• Ex. Pointing Modes not Control Accuracy
• Dont specify characteristics of
  hardware/software that has already been baselined
• Less than 100 requirements for entire Roadrunner
  bus
• Makes for quicker verification test planning
• Keeps focus on the key issues that you can affect

6

System Design Keys

• Stick to whats available immediately or sooner
• Parts built from in-house stock
• Spares from other flight programs
• COTS HW with short delivery times
• Dont optimize, dont change
• The first tweak costs a disproportionately
  large amount of time and money
• Resist pressure to squeeze more utility by making
  a minor adjustment you cant build until you
  freeze the design
• Use whole design solutions
• A whole intact set eliminates the need for new
  interfaces
• Use automated design tools

7

Excellence Aint Always Pretty

• Scrounging and Kluging may make the marketing
  types cringe, but the soul of an elegant
  optimization for schedule often means making do
  with whats at hand
• Get over your biases and focus on what works best
• Although it isnt custom designed for your
  application, there is a distinct challenge in the
  art of choosing the best existing hardware to
  make a system work
• New interfaces can be a killer, try to choose
  parts that work well together with minimal
  hardware changes
• Mixing pedigrees of parts is OK, but dont exceed
  the environments unless you can afford the
  testing
• Quality comes from a good design and adequate
  testing
• Where its from doesnt matter as long as it
  works together

8

Documentation

• Minimize, minimize, minimize
• Keep the core essentials ICDs, Block Diagrams,
  Schematics, Drawings, Test Plans/Results, Budgets
• Minimizing writing/reading frees up more time to
  talk face to face and work the real issues
• Shorter schedules tighter focus make it less
  likely things will be forgotten
• Can rely more on informal documentation such as
  E-mail notes providing that all key personnel
  participate
• Centralized repository with universal access
  speeds coordination and upkeep of documents

9

Roadrunner Example
10

Communication/Organization

• Small teams are much more efficient at
  communication
• Requires the right people smart, proactive,
  communicators
• Empower team members to make decisions
• Train them how and when to coordinate those
  decisions with others to keep the design
  consistent
• Single organization, co-located is best,
  otherwise
• Structure contracts to allow technical solutions
  to be resolved dynamically without the need to
  renegotiate statements of work at each turn
• Incentives for reaching mission objectives, not
  organizational deliverables
• Keep SOW general and flexible enough to adapt as
  design matures, trust mission objective
  incentives to keep players honest

11

Conclusions

• Optimizing for responsive design is not just
  about cutting schedule or adding more design
  resources, a different approach is required
• Setting appropriate requirements and synthesizing
  solutions from a restricted set of available
  components is much more important that skill in
  customized design
• Keeping the mission capabilities driven is a key
  enabler
• Minimized documentation and small, empowered
  teams allow the process to be executed efficiently