Title: CSCI 308 Systems Analysis
1Systems AnalysisCSCI 308
- Instructor Stan Schuyler
- Lecture 6
- Structured Problem Solving Approaches
- SA Determining the Real Problem
- (Lecture 5 cont)
2Topic Outline
- Recap of Team Project 0
- Former SAD Student Inputs Matt Machuga
- What is the Problem?
- Description of a Structured Analysis Technique
- New Team Assignments
- Quiz
3What Does an SA Have to Produce?
- Work Products
- Symptom Analysis
- Problem (Causation) Statements
- Solution Proposals
- Proposed Architecture
- Business Cases
- Project Plans
- What Table of Contents do each of the above have?
4Questions Find ltInterpretationsgt
- Key Questions
- What is produced?
- Why produced?
- Who or what is in action?
- Where is the action occurring?
- How is production done?
- When is the action occurring?
- What is the data?
- How represented?
- How presented?
- What is the information?
- What is the knowledge?
- What level of Quality Control?
- Find ltinterpretationgt
- (Artifacts, Performance)
- (Problem, Purpose)
- (Actor/Agent)
- (Place, Situation/context)
- (Methodology)
- (Time, Sequence)
- (Symbols, Language)
- (alphabet, grammar)
- (Conveyed, transmitted)
- (Meaning in symbols)
- (Why How of meaning)
- (Feedback)
5What is the Problem?
- What is problem recognition and problem
definition? - Explain the differences between problems and
symptoms - What is
- A PIECES framework?
- An Ishikawa diagram?
- An EATPUT diagram?
6The Biggest Problem!
- What you think you know!
- What you do based on habit because of what you
think you know!
7Problem Recognition and Definition
- Cause and Effect
- One common method to define a problem is to
explore the cause and effect trail created by its
symptoms. - If one can discover a common source for the
symptoms, then one can assume that he or she has
determined the nature of the problem. - But how do we specify How to do this?
8A Look at Causation Categories
- Performance
- The symptom indicates some problem related to the
systems response time, response rate, or
availability. - Information
- The symptom indicates a requirements
specification deficiency, error, or
interpretation mismatch. - Economics
- The symptom indicates a competitiveness problem
in the sense that the system (personnel and
application) is too expensive to operate relative
to competitors operation. - The reference is internal to external
9Causation Categories (cont)
- Control
- The symptom indicates a problem with the
repeatability, reliability, stability, or control
of a process. - The symptom could be indicative of a security
issue in the sense of process integrity The
user is feels the process is protected from
failure. - Efficiency
- The symptom indicates that there is an overuse of
resources that limits capacity in some way (e.g.
too much CPU limiting the throughput of a
process). - The issue is internal to the system
10Causation Categories (cont)
- Service
- The symptom indicates the output is incorrect,
undesired or missing. - In this sense service relates to what is supposed
to be produced for the user outcomes.
11PIECES Natural Interdependencies
Performance Causes -delays -excessive
effort/ time
Office/Decision Control Causes -Mgmt. Culture
-Processes gtadhoc gtunchecked gtover
managed gtdecisions gtethics -Security
gtdata gtaccess
Information (requirements) Causes -missing
-in error -noise -unavailable
SYMPTOMS OBSERVED
Efficiency Causes -Waste gttime
gtmaterials gteffort -Redundancies gtData
gtTasks
Cause
Cause Because
Cause Because
Cause Because
12(No Transcript)
13Problem Causation
- Design Microchip Engineering Dept. GRED Support
Report - Addressed to H. Gummel, Dept. Head, Computer
Aided Design Dept. - From Dick Jacobs, Dept. Head, Microchip
Engineering Dept. - This report describes a history of problems
we have encountered since the first release of
your GRED XYMASK chip layout editor. First, let
me make it clear that GRED has proved to be an
excellent tool for custom chip design. That said
we have experienced several recurring problems
that we need a solution to. Let me briefly
summarize the history of events.
14What do we need to Extract?
- WHO (sources)
- Players
- Roles
- What (Symptoms) facts, not inferences
- When (sequence ? patterns)
15Problem Analysis
- Who
- H. Gummel, Dept. Head, Computer Aided Design
Dept. - Role Problem Solver
- Dick Jacobs, Dept. Head, Microchip Engineering
Dept. - Role Client and Customer Representative
- This report describes a history of problems
we have encountered since the first release of
your GRED XYMASK chip layout editor. First, let
me make it clear that GRED has proved to be an
excellent tool for custom chip design. That said
we have experienced several recurring problems
that we need a solution to. Let me briefly
summarize the history of events.
16More What
- 5/19/81 GRED 1.0 Released
- 5 Circuits were designed and 5 XYMASK layouts
produced. These were sent to the foundry for
producing test wafers at 20,000 each.
Examination of the test wafers indicated that
certain types of geometries were not positioned
relative to one-another correctly due to a
misunderstanding between the users and your
software team. We issued MR1 through MR3 to
request specific fixes.
17Example of Analyzing the ME Dept. Doc.
18Another Framework
- There is no universally correct framework for
analysis - You will adapt one you see or invent your own
when presented with a new domain. - Another possibility is to categorize based on
- Symptom (WHAT)
- Who
- When
- Where
- Outputs
- Inputs
- Data Representation
- Summary of Apparent Cause
19(No Transcript)
20Team Assignments
- Team Re-formation based on
- Anonymous Leadership Selection
- Public Round Robin nomination
- Call out R number
- Member with R number may choose to join or
- Member with R number may remain anonymous and
not identify themselves (unless they are the last
one ?) - Assignments
- Do not forget Assignments 6A and 6B
- Assign 7B
- Assign 7C
21Quick Paper Quiz
- Put your name and date on it
- Question the answers!! ?
- And Answer the Questions ?