systems textbook GSGS

system

set of elements so interconnected as to aid in driving toward a defined goal

• objects w characteristics in common

• objects interconnected/influence each other

• elements achieve defined goal/objective

IP

how good is system/how close it is to achieving goal

system has _____ objective meaning

no

RAND approach

systems analysis = OR + PA

operations research , policy analysis

analysis

taken apart

design

synthesis

systems analysis is

BROAD

SE = SA

systems engineering is systems analysis which basically is systems integration

operations research

analytival ways with optimization for client

policy analysis

management/goals of issue

taylorism

rigid separation of managers’ thinking & workers’ doing

• hierarchal

large scale system (LSS)

• policy component - effectiveness evaluated by higher order judgements by simple economic efficiency

• high order - larger number of parts

• complex to describe

• lengthy installation

• unique

• prior complete testing impractical - b/c of size & cost, impractical to construct a test prototype before installing OS

intelligent transportation systems (ITS)

use of disparate technology to improve (without greater capacity) performace of transformation

• high order system

• demographics < policial questions < sociological factors - contribute to complexity of system, difficult to predict

• each is unique

• components and subsystems must be carefully field-tested prior to final installation

• care for system > component in analysis

systems integration

logical, objective procedure for applying in an efficient, timely manner new and/or expanded performance requirements to design

what defines a systems study

1. top down nature (specific goals)

2. goal-centered approach

3. rational, objective basis for analysis

4. analytic/quantitative + policy

5. generalized problem (including problem setting) → humans

6. optimization (analytical modeling & stimulation)

7. explicit analysis of operative values assumed & declaration of analysts’ biases and interests

8. problem/client orientation rather than technique/abstract orientation

10 golden rules of systems analysis

1. always a client

2. client does not understand his own problem

3. must generalize problem to give it contextual integrity

4. client doesn’t understand the concept of the IP

5. you are analyst, not decision maker

6. meet time deadline and cost budget

7. use good-centered not technology of chronological approach

8. non-users must be considered in analysis & final recommendation

9. technology

10. decision marker is confused one

systems study

1. determine goals

2. establish criteria for ranking alternative candidates

3. develop alternative solutions

4. rank alternative solutions

5. iterate

6. action

non performance concerns

• effect on nonusers

• impact on existing system

• ratification procedures

• effect of incremental introduction

• sensitivity of IP to parameter variation

• emergent systems

how many times should we iterate?

should consume smoothly to optimal result

each iteration clarifies & solution best costs time & money

stop when: additional cost of one more iteration > probable benefit of additional clarification

systems analysis

achieve elements’ goal

goal centered rather than technologically centered approach

transition scenario

step by step, organized procedure for achieving goals

top down planning vs bottom up planning

general to specific, based on goals & objectives

based on current conditions, uses current technology & minor extrapolations, employs incremental, step by step approach

strengths/weaknesses of top down planning

strengths: based on assumptions & trends of planning environment, adjusts to environment

weaknesses: lose focus & objectivity if not executed properly, does not provide sufficient debris for effective action on short range issues

strengths/weaknesses of bottom up planning

strengths:

• lends itself to immediate evaluation of cost effectioness

• consistent with conventional engineering design methodology

weaknesses:

• focus vision on short term problems

• produces ever decreasing incremental improvements

• locks in current technology & operational structure

top down analysts

corrects errors of bottom up approach, tries to stop surrounding world

descriptive scenario

narrative that describes status quo

normative scenario

narrative that describes problem environment when ideal system is inplace & opening successfully

transition scenario

here (DS) to there (NS)

• graph chart with major project steps & milestones

• delineations of constraints, critical matrices - interaction matrices are effective in clarifying interactions

• cost of failure to stay on schedule

• responsibilities of major actors

• sign offs required of major cooperative groups

___centered, not _____ centered

goal, technology

conventional engineering is

bottom up

4 possible goals for large corporation

1. maintain share of market while maximizing (long term) ROI

2. maximize sales

3. maximize (short term) profits

4. maximize market control

how does iteration promote efficiency?

1. helps get study started

2. lower fear of making initial earns

3. lower wasted effort on unnecessary subtasks/extraneous dead ends

life cycle of a system

• problem analysis

• system design

• system construction and installation

• acceptance testing and operation

• maintenance and periodic upgrading

• decommissioning, dismantling, replacement

share of life cycle cost & life cycle phase

3% - establish scope of project

9 - 20% - requirements specification

8% - system design

10 % - coding

15 - 30% interaction testing

5% - acceptance testing

50% - operations, maintenance, version upgrades

goal development

costs little, but if he not done correctly, can cost the whole project

most sensitive

1. generalize the question

2. develop a descriptive scenario (status quo)

3. develop a normative scenario (fully operative) → change in features, presume features

4. develop the axiological component (values)

5. prepare objectives tree (goods) - hierarchal, graph thing - to (action) + (object) + (qualifying words)

6. validate (evaluation)

7. iterate

inscoping

specific, detailed

outscoping

broader view

diagram

each goal statement should provide greater explicit/detailed goal than statement above it

reading up - each higher statement answers why goal below it is needed

reading across - are all these more specific goals needed to accomplish more general goal?

• what other specific objectives are needed to accomplish the more general objective?

 ideal Index of Performance has the following five characteristics

Measurable Objective Nonrelativistic Meaningful Understandable

compound interest

F = P[(1 + i)n ] 

 value of constant annual payments

A = P[i(1 + i)n /((1 + i)n − 1)]

four common criteria of economic efficiency

1. internal rate of return

2. benefit-cost ratio

3. present net worth

4. payback period

what is wrong with b-c ratio

• nothing for distributive justice

• individual cash items can be interpreted as either an additional benefit or a reduced cost, and the converse is also true

• can incorrectly optimize the per unit investment

IRR ___

can be fixed

expected monetary value (EMV) 

“the monetary value of the event, if it occurs, times the probability of occurrence.

bottom-up approach

flawed because 1. Work with implicit goals as embodied by the client in concrete design specifi-cations.2. Ignore potential conflicts until and unless they occur, and then handle on an adhoc basis.3. Work with monetary criteria—that is, to minimize cost and/or maximize revenue.

causal loop diagrams

things affect each other, circle cycle

R - reinforcement (away from equilibrium)

B - balance (toward equilibrium)

reference behavior period

graphical rep of behavior over time of >1 variables in loops we are analyzing

observed behavior period: show historical states of variables at a given period

observed behavior period

show historical states of variables at a given period

\\ in graph

delay

system: different definitions

• objectives

• environment

• resources

• components

• management

measuring performance of component is

tricky

stated objective vs. real objective

stated: need to use more precise measurements of performance in overall system

systems thinking

must consider wider implications

environment

out of system’s control & determines in part how system performs

management scientist

what can i do about what i have?

• careful analysis

• missions rather than depts

• anti-political

• receives info on whether or not erroneous & must include steps that provide for change

technological advances →

more resources

cybernetic loop:

various info systems that inform → possibly change plan

stock

foundation of system

flow

inflow/outflow, change

diagram of inflow, stock, outflow

inflow → stock → outflow

dynamics

behavior of stocks & flows over time

dynamic equilibrium

inflow = outflow no change

stock increase if

stock decrease if

inflows > outflows

inflows < outflows

human minds focus more on

stocks > flows

human minds focus more on

inflows > outflows

stocks can be

delays/buffers

change in stocks →

dynamics

stocks

allows inflows/outflows to be independent of each other & temporarily out of balance with each other decisions → regulate stocks

systems thinkers

see world as collection of “feedback processes”

feedback loop

closed chain of causal connections from stock (2 directions), dependent on load of stock

systems diagram

simplifications of real world

balancing feedback loop

goal/stability seeking

feedback loops

1. goal seeking

2. amplifying, reinforcing, self multiplying (reinforcing feedback loop)

Feedback Loop (set phasers to sun)

The process by which a system informs the user of its current state; in the Therac-25, this loop was broken or misleading.