Tradeoffs - Applying Systems Engineering - Lecture Slides, Slides of Systems Engineering

Download Tradeoffs - Applying Systems Engineering - Lecture Slides and more Slides Systems Engineering in PDF only on Docsity!

Architecture Selection - Trade

Studies

Objectives

• Identify Key Terminology

• Describe approach to diverseecture selection

• Describe a generic Trade Study Methodology

• Apply Trade Studies Methodology to conduct

Functional Trades.

The Pitfall of the Point Solution

Key Terminology

• Architecture

• diverse Consequences Analysis

• Decision Criteria

• Functional Alternatives

• Killer Trades

• Sensitivity Analysis

• Trade Study

• Trade Tree

Architecture –

The Problem

• And, just how do we do this?
• First step in Architecture Definition: Identify potential

solutions.

• Hard work, creativity, inventiveness, research and development,

thoroughness, intelligence, historical successes, methodical

approaches, analysis, modeling, experimentation, muddling, brute

force, and so on provides alternatives.

• Selection of the “Best of Show”
  • Trade studies or decision analysis is then used to select the best

available architecture (solution).

• And what is Best?
  • Therein lies the rub.

Identification of Possible Architecture

• Pick last year’s solution
  • It worked before, it can work again. Classic approach: “Do it the way we’ve always done it!”
• Chief Engineer
  • Refer to the company expert
• Use what is available
  • Take what you have and force it to work.
• Muddling
  • Try everything you can think of or afford. Work without a plan, don’t keep records, you won’t know what went wrong before.
• Try everything
  • See solution of typical mind problem, Excel file “Houses and Colors”. Read Problem first, on sheet marked “Problem.” Then see Solution sheet.
• Systems Engineering
  • Systematically review functions and requirements; convene integrated product teams, hold braining storming sessions, hold Value Engineering sessions, generate a multitude of ideas in open work sessions. Repeat if necessary
• Model Based Design
  • Convert Functional Flow Block Diagram into a model (computer, mathematical, or mechanical). Identify a unit operation to perform each function. Continually test, update and improve model. Example – Model of SBW treatment.

Selection of the “Best of Show”

• We now have a number of candidate solutions, ranging from 0

to infinity

• Next step is to select the “preferred choice”

• The preferred choice should be the “best choice”

• But what is the Best?

• Cheapest, fastest, highest quality, optimum, CEO’s choice, default if only one is available,

• Decision analysis is then used to select the best available

architecture (solution)

• If there is 0 (no solution) or 1 (point solution) candidate, the

decision is trivial

• If there are more than 1 candidate, then the analysis

performed to support the decision is a Trade Study

About the Term

Trade Study

• Often referred to as Trade-off Study

• Incorrectly written Tradeoff Study

• Google search shows that Trade Study is preferred over Trade-

off Study 5 to 1

• ISO 15288 uses Trade-off Study, INCOSE Handbook uses Trade Study.

• Often used interchangeably with alternatives study

• INEEL Engineering just does a trade and publishes it in an EDF

(Engineering Design File) without a name.

• Value Engineering is a serious contender, often confused with

a Trade Study

Why do Trade Studies?

• Management cannot make key programmatic decisions without detailed,

rigorous, verifiable data.

• Customer expects/demands trade studies be performed on all major

system elements.

• Need something to provide technical defensibility for the program
• Trade studies provide an accepted method to document choices for

management decisions.

• We do trade studies automatically in the course of good engineering

practice.

• Even if a trade study reveals that the point solution (single candidate) was

the preferred solution, it now has documented justification as the

preferred solution.

Trade Studies Consist of Many

Facets

Decision Trees Baseline Configuration

Data Sheets Analyses Simulations

Objectives

A B C D A B C D E^ E

Candidates

Wt

Criteria

COMPARATIVE

DATA

Complete Trades Documentation and Subject to CM

Identify System Decisions

Develop Decision Criteria based on Req’ts

Develop Suitable Alternatives

Weight Decision Criteria

Score Alternatives

Compute Total Weighted Scores and Eliminate (Capture in Documentation)Rankings Insensitive Criteria Perform Sensitivity Analysis

Perform Adverse Consequences Analysis

Eliminate Unsuitable Alternatives

Perform Killer Trades

Develop Trade Tree of System 2 Decisions^1

3

4

(^65)

7

8

9

11

10

9a

10a

Trade Study Approach

Steps 1-5 (This is at program start!)

1. Determine what System Design Decisions need to be made
2. Develop a Trade Tree of System Decisions based on current knowledge
3. Establish System Design Decision Criteria.

Stakeholders and decision makers should concur with these criteria. SE’s structure identification. Often times, Stakeholders don’t care until they see potential results. However, CERCLA has formally established criteria a) Based on System Requirements b) Nmerical if possible c) Threshold (must have) and Balancing Criteria (desirable attributes) d) Independent

4. Assign weights to the Decision Criteria
5. Develop all suitable Alternatives

Complete Trades Documentation and Subject to CM

Identify System Decisions

Develop Decision Criteria based on Req’ts

Develop Suitable Alternatives

Weight Decision Criteria

Score Alternatives

Compute Total Weighted Scores and Eliminate (Capture in Documentation)Rankings Insensitive Criteria Perform Sensitivity Analysis

Perform Adverse Consequences Analysis

Eliminate Unsuitable Alternatives

Perform Killer Trades

Develop Trade Tree of System 2 Decisions^1

3

4

(^65)

7

8

9

11

10

9a

10a

Trade Study Approach

Steps 6-

6. Perform a Killer Trade if needed (Killer is a

jargon term used by some to mean trade criteria that eliminate the alternative on a single criterion, also known as Musts vs Wants in KT.)

7. Have discipline experts score the

alternatives

8. Capture trade study results on a Decision

Analysis Worksheet or other form

9. Perform a Sensitivity Analysis. Does any

one variable or combination of variables

drive the decision? Show on the Analysis

Worksheet.

10. Perform an Adverse Consequences

Analysis. Revise the alternatives if

needed.

11. Final documentation of trade. (Should be

placed under configuration control.)

System (^) Top-level

Functions

Alternative Architectures

Step 2: Develop a

Trade Tree

Electric Power System

Store Energy

Control & Distribute Power

Generate Power

Solar Array Eclipse/Peak

Nuclear Load Leveling

Autonomy

Circuit Protection

Bus. Reg.

Nuclear

Solar Array

NiCd

NaS

NiH

NiH

NiCd

S/C Data Processor

EPS Microprocessor

Circuit Breakers

Fuses

120 VDC

28 VDC

Rotary

Thermionic

Flexible

Rigid

Regulated Dual Bus

Unregulated

GaAS

Si

Unregulated

Step 3: Establish

Decision Criteria

• How do we select between alternatives?
• Look at:
  • System Requirements
  • TPMs
  • compliance measures
  • other factors (i.e., political issues, public concerns)
• Determine how to measure them
  • absolute vs. comparison
  • units or scales
• Focus on the discriminators — you won’t have the resources to study

everything

• Always Consider Cost and Schedule Impacts
• Document Them