Information Engineering

1
Information Engineering
Enhancing Information Quality for Effective
Decision-Making in Textiles
for
The 2000 Massachusetts Institute of
Technology Conference on Information Quality
October 21, 2000 Cambridge, Massachusetts
Presented By Mr. Neil Cahill Vice
President Institute of Textile Technology Charlott
esville, Virginia with
Dr. George Hodge North Carolina State University
Dr. William Oxenham North Carolina State
University
Mr. Yatin Karpe North Carolina State University
i\share\sandys\conferences\informationengineering
-mit2.ppt
2
. . . plant having Performance Capability a
traditional plant does not have, and, therefore,
cannot match performance, no matter how modern
or well managed.
i\share\sandys\conferences\informationengineering
-mit2.ppt
3
ONE NEW CAPABILITY NEEDED
. . . High proportion of current manufacturing
performance losses are due to slow response to
production problems.
i\share\sandys\conferences\informationengineering
-mit2.ppt
4
PERFORMANCE DEGRADATION CURVE
Point of Degradation
X
Normal
Point of Malfunction
Process
X
Performance
Point of Failure
X
Poor
Response Delay Time
i\share\sandys\conferences\informationengineering
-mit2.ppt
5
SLOW RESPONSE CONTROL
Good
X
Performance Loss
X
Performance
Measures
Reaction Point
X
Poor
Short
Long
Response Delay Time
i\share\sandys\conferences\informationengineering
-mit2.ppt
6
FAST RESPONSE CONTROL
Good
X
Loss
Reaction Point
X
Performance
Measures
X
Poor
Short
Long
Response Delay Time
i\share\sandys\conferences\informationengineering
-mit2.ppt
7
TIME VALUE OF INFORMATION
i\share\sandys\conferences\informationengineering
-mit2.ppt
8
. . . Value of Information decays rapidly with
elapse time after need to know.
i\share\sandys\conferences\informationengineering
-mit2.ppt
9
TIME VALUE OF INFORMATION
Use Value of Information
Need to Know
TIME
Info Available
i\share\sandys\conferences\informationengineering
-mit2.ppt
10
. . . Finding and Diagnosing information takes
80 of total time needed to complete a decision!
i\share\sandys\conferences\informationengineering
-mit2.ppt
11
DECISION RESPONSE TIME
Decision
Action
Slow Reaction
Find
Diagnose

10
10
80
100
Fast Reaction
Fd
Diag
Dec
Act
10
10
10
5
35
i\share\sandys\conferences\informationengineering
-mit2.ppt
12
Measuring Decision Effectiveness
(Manufacturing Problem Solving)
i\share\sandys\conferences\informationengineering
-mit2.ppt
13
BASIC DETERMINANTS OF MANUFACTURING DECISION
EFFECTIVENESS
1. Number of process malfunctions. 2. Delay in
responding to problems. 3. Effectiveness of
corrective action taken.
i\share\sandys\conferences\informationengineering
-mit2.ppt
14
DECISION EFFECTIVENESS EQUATION
(
(
)
(
)
)
(
)
(
)

Malfunction Frequency
Response Delay
Defects/ Hr
Defects/ Hr
Elapse Time

Decision Effectiveness
1.0-R.E.
Failure Rate
Reaction Delay
Action Effectiveness
R.E. Response Effectiveness
15
TWO CONCEPTS OF AN INFORMATION SYSTEM
i\share\sandys\conferences\informationengineering
-mit2.ppt
16
INFORMATION DELIVERY SYSTEMS
1. Information Distribution System. 2. Decision
Message System.
i\share\sandys\conferences\informationengineering
-mit2.ppt
17
Users
Information Flow Pipe
General Report
Traditional Information System
i\share\sandys\conferences\informationengineering
-mit2.ppt
18
WHAT DO USERS REALLY WANT?
Information in Reports? OR Meaning in
Information?
i\share\sandys\conferences\informationengineering
-mit2.ppt
19
. . . it is Meaning which allows people to
understand situations and make appropriate
decisions
i\share\sandys\conferences\informationengineering
-mit2.ppt
20
MESSAGES
. . . Intended meaning to be conveyed by the
information to the user.
i\share\sandys\conferences\informationengineering
-mit2.ppt
21
. . . I told him 10 times (Information), but he
still doesnt understand whats happening
(Message)!
i\share\sandys\conferences\informationengineering
-mit2.ppt
22
Users
Customized Report
Expected Decision
Message Interface
Information Flow Pipe
Decision Message System
i\share\sandys\conferences\informationengineering
-mit2.ppt
23
Technician
Manager
Customer Message Interface
Information Flow
Customized Message Interface For User
i\share\sandys\conferences\informationengineering
-mit2.ppt
24
INFORMATION ENGINEERING
. . . method to extract the Meaning Content from
information before sending to user.
i\share\sandys\conferences\informationengineering
-mit2.ppt
25
. . . information needs are Decision-Driven
i\share\sandys\conferences\informationengineering
-mit2.ppt
26
DATA TO DECISION CYCLE
. . . sense an event
Detection
. . . identify and quantify event
Data
. . . relationship of different events
Information
. . . meaning of relationships
Diagnosis
. . . understanding relationship
Message
. . . choice of response
Decision
. . . implement change
Action
i\share\sandys\conferences\informationengineering
-mit2.ppt
27
DATA TO DECISION CYCLE
Detection
Data
Information Reporting
Information
Diagnosis
Message
Decision
Action
i\share\sandys\conferences\informationengineering
-mit2.ppt
28
DATA/INFORMATION REPORT
Weaving Computer Monitoring Report
MCH
EFF
STY
WVR
S/RLH
WS
FS
OS
0049 0052 0051 0053 0054 0055 0037 0038 0039 0040
0561 0562 0563 0564 0565 0001 0002 0003 0004 0005
0572 0573 Avg.
091.0 080.4 083.1 074.1 091.6 091.6 105.2 094.1 09
4.5 086.0 094.6 096.6 098.0 094.0 094.1 072.1 086.
8 089.5 079.5 094.1 094.8 095.0 090.0.
056 056 056 056 056 056 056 056 053 056 056 053 05
6 056 053 056 056 053 056 053 056 053
032 033 033 033 033 032 032 032 033 033 033 032 03
2 033 032 033 033 032 033 032 032 032
1.09 1.24 0.90 2.02 0.27 1.22 0.00 0.13 0.13 0.72
0.26 0.65 0.38 0.13 0.53 2.08 0.72 1.68 0.94 0.53
1.18 0.92 0.80
5 6 5 11 2 8 0 1 1 3 2 3 3 1 3 4
1 3 2 1 2 5
2 2 0 1 0 1 0 0 0 2 0 2 0 0 1 8
3 7 4 3 1 2
1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0
1 2 0 0 6 0
i\share\sandys\conferences\informationengineering
-mit2.ppt
29
DATA TO DECISION CYCLE
Detection
Data
Information
Diagnosis
Meaningful Messages
Message
Decision
Action
i\share\sandys\conferences\informationengineering
-mit2.ppt
30
MEANING/MESSAGE REPORT
o
100
o
o
o
o
o
Weaver 032
o
95
o
o
o
o
90
Machine Efficiency ()
85
80
Weaver 033
75
0
2
4
6
8
10
12
14
Stops Rate
Efficiency Delay Curve
i\share\sandys\conferences\informationengineering
-mit2.ppt
31
Users
Customized Report
Expected Decision
Message Interface
Information Flow Pipe
Decision Message System
i\share\sandys\conferences\informationengineering
-mit2.ppt
32
Walk-through Visit of 21st Century Textile
Plant
i\share\sandys\conferences\informationengineering
-mit2.ppt
33
CIM
Computer-Integrated Manufacturing
i\share\sandys\conferences\informationengineering
-mit2.ppt
34
Figure 1. Automated and computerized plant with
full-factory communications network (Murata CIM
Mill).
i\share\sandys\conferences\informationengineering
-mit2.ppt
35
. . . Future textile plants will be highly
automated!
i\share\sandys\conferences\informationengineering
-mit2.ppt
36
Automated Piece Up and Cone Doffing
Academic\No5Class899.ppt
38
i\share\sandys\conferences\informationengineering
-mit2.ppt
37
45
i\share\sandys\conferences\informationengineering
-mit2.ppt
38
PLANT COMMUNICATIONS NETWORK
. . . Local Area Network (LAN) providing real-time
bi-directional communication between individual
machines and plant computer.
i\share\sandys\conferences\informationengineering
-mit2.ppt
39
COMMUNICATIONS CENTER
Assess Plant Performance Status and Broadcast
Action Messages
i\share\sandys\conferences\informationengineering
-mit2.ppt
40
i\share\sandys\conferences\informationengineering
-mit2.ppt
41
Broadcast Messages To Selected Personnel
O
Computer Control Center
O
Broadcast
Message
O
Factory Network
Machine-to-Man messaging
i\share\sandys\conferences\informationengineering
-mit2.ppt
42
SIGNAL DISPATCHED OPERATORS
. . . Computerized monitoring system detects
Machine Service Needs and automatically Calls
appropriate personnel.
i\share\sandys\conferences\informationengineering
-mit2.ppt
43
O
O
Technician
Operator
O
O
Handler
Cleaner
Spinning Machine Change-Over Swat Team
i\share\sandys\conferences\informationengineering
-mit2.ppt
44
So now must . . .
Put intelligence into the machine system!
i\share\sandys\conferences\informationengineering
-mit2.ppt
45
i\share\sandys\conferences\informationengineering
-mit2.ppt
46
AUTOMATED DECISION OF FABRIC GRADE
IF OR THEN IF NOT
More than (3) Stops/50 yards Technician Repair
Required Send AGV to Human Inspector Send AGV
to Loading Dock
i\share\sandys\conferences\informationengineering
-mit2.ppt
47
ENGINEERING INFORMATION FOR MANUFACTURING
DECISION MAKERS
i\share\sandys\conferences\informationengineering
-mit2.ppt
48
INSTANTANEOUS FACT-FINDING
. . . provide a decision-maker with needed
facts in the logical order to readily understand
performance problem as soon as detected.
DECISION MAKER DRAWS THE CONCLUSIONS
i\share\sandys\conferences\informationengineering
-mit2.ppt
49
TYPES OF DECISIONS MADE IN MANUFACTURING
i\share\sandys\conferences\informationengineering
-mit2.ppt
50
CLASSIFY DECISION TYPES
1. PROGRAMMABLE DECISION
. . . repetitive situations with predictable
response.
2. DISCRETIONARY DECISION
. . . Situation requiring a judgment call.
3. KNOWLEDGE INTENSIVE DECISION
. . . Unique situations requiring special
knowledge for response.
i\share\sandys\conferences\informationengineering
-mit2.ppt
51
TYPE DECISIONS MADE IN MANUFACTURING
100
80
60
Manufacturing Decisions
40
20
0
Programmable
Discretionary
Knowledge Intensive
i\share\sandys\conferences\informationengineering
-mit2.ppt
52
. . . concentrates on Programmable
Decisions avoiding Discretionary or
Knowledge-Intensive Decisions, which are highly
complex and uncertain.
i\share\sandys\conferences\informationengineering
-mit2.ppt
53
DESIGNING THE MESSAGE INTERFACE
i\share\sandys\conferences\informationengineering
-mit2.ppt
54
MESSAGE INTERFACE
. . . Part of communications system that converts
information into Meaningful Message so user can
make quicker and more effective decisions.
i\share\sandys\conferences\informationengineering
-mit2.ppt
55
Technician
Manager
Customer Message Interface
Information Flow
Customized Message Interface For User
i\share\sandys\conferences\informationengineering
-mit2.ppt
56
DATA TO DECISION CYCLE
Detection
Data
Information
Diagnosis
Message Interface to User
Message
Decision
Action
i\share\sandys\conferences\informationengineering
-mit2.ppt
57
DESIGNING THE MESSAGE INTERFACE
1. Identify Machinery Layout. 2. Identify
Information Sources 3. Identify Information
User 4. Determine Expected Decision. 5.
Determine Need-to-Know Information. 6. Develop
Diagnostic Logic. 7. Design Actionable
Message Format. 8. Assess Action
Effectiveness.
i\share\sandys\conferences\informationengineering
-mit2.ppt
58
STEP 1. MACHINERY LAYOUT
. . . specific type and arrangement of production
machinery to be controlled.
i\share\sandys\conferences\informationengineering
-mit2.ppt
59
TYPE MONITOR
PROCESS NAME
MACHINE LAYOUT
Barco Sycotex
Stretch Breaker
Rebreaker
Barco Sycotex
Blender
Servo
Sliverdance
Roving
Barco Sycotex
FA.NI Monitor
Ring Spinning
Savio Elios
Linked Winding
Loepfe Clearers
Barco Sycotex
PACKING
i\share\sandys\conferences\informationengineering
-mit2.ppt
60
STEP 2. IDENTIFY INFORMATION SOURCES
. . . Determine what computer accessible sources
of information are available.
i\share\sandys\conferences\informationengineering
-mit2.ppt
61
INFORMATION SOURCE DIAGRAM
. . . Identifies types of information
and monitoring source at each manufacturing proces
s.
i\share\sandys\conferences\informationengineering
-mit2.ppt
62
INFORMATION SOURCE DIAGRAM
Process
Information Type
Stretch Breaker
Production
Output Pounds
Pounds/Hour
Efficiency
Stops
Declaration (Long Stops)
Maintenance
Service
Stop Cause
Rebreaker
Production
Output Pounds
Pounds/Hour
Efficiency
Stops
Declaration (Long Stops)
Maintenance
Service
i\share\sandys\conferences\informationengineering
-mit2.ppt
Stop Cause
63
STEP 3. IDENTIFY INFORMATION USER
. . . Specific individuals, technicians,
operators, managers, etc., who make decisions in
response to problem situations.
i\share\sandys\conferences\informationengineering
-mit2.ppt
64
. . . Target User is maintenance technician.
i\share\sandys\conferences\informationengineering
-mit2.ppt
65
STEP 4. DETERMINE EXPECTED DECISIONS
. . . decisions routinely made by
Individual responsible for responding to a
particular problem.
i\share\sandys\conferences\informationengineering
-mit2.ppt
66
MAINTENANCE TECHNICIAN EXPECTED DECISION
PROBLEM SITUATION
EXPECTED DECISION
1. High Flag Machines
Quickly identify High Flag Machines and take
corrective action with high First Time Success
rate.
2. High Off-Quality Machines
Identify quality defects while On-Machine and
correct immediately.
3. Meet Maintenance Schedule
Meet all maintenance schedules and assure
adequate parts inventory.
i\share\sandys\conferences\informationengineering
-mit2.ppt
67
STEP 5. DETERMINE NEED-TO-KNOW INFORMATION
. . . Specific information in logical order
needed by decision-maker to understand situation
and make right choice.
i\share\sandys\conferences\informationengineering
-mit2.ppt
68
MAINTENANCE TECHNICIAN Need-To-Know
High Flag Machines
1. Which Machine? 2. What Style? 3. How
Long? 4. Type of Flags? 5. Flag History.
6. Maintenance Record. 7. Parts Replacement
Records 8. Assigned Technicians. 9.
Assigned Weaver. 10. Any Special Events
i\share\sandys\conferences\informationengineering
-mit2.ppt
69
STEP 6. DEVELOP DIAGNOSTIC LOGIC
. . . Arrange need-to-know information into the
logical order of inquiry that a decision-maker
would follow to understand what happened and how
to respond.
i\share\sandys\conferences\informationengineering
-mit2.ppt
70
TWO TYPES OF INQUIRY MADE BY COMPUTER
QUESTIONS (Qs) . . . information in
database go get it!
MESSAGES (Ms) . . . information not
available in database ask people to find it!
i\share\sandys\conferences\informationengineering
-mit2.ppt
71
COMPUTING DIAGNOSTIC LOGIC
High Flag Machine
Diagnostic Stage
Order of Inquiry
Question/Message Statement
I. Problem Focus . . . What problem to
diagnose?
Q1
Who is the User?
What Malfunction triggered alarm?
Q2
Which Machine triggered alarm?
Q3
What Product being run?
Q4
II. Localize Problem . . . Where is
problem located?
Q5
How long has problem been happening?
All or some Machines?
Q6
All or some positions on machine?
Q7
Q8
Any other alarms on this machine?
M1
Message to Maintenance Check settings on alarm
machines
i\share\sandys\conferences\informationengineering
-mit2.ppt
72
STEP 7. DESIGN ACTIONABLE MEASURE FORMAT
. . . Did report to user (1) provide
needed information, (2) explain what happened,
and (3) allow an effective decision?
i\share\sandys\conferences\informationengineering
-mit2.ppt
73
ALERT/ADVISORY MESSAGE
TO
Bob Roberts, Maintenance Technician
ALERT
Machine 18 classified as Malfunctioning with
Chronic High Flags.
ADVISORY
Over past 5 days (1/1-1/5), 65 of checks
out-of-limit, including stops, efficiency, flags,
and quality points.
ACTION
Flags which average 2/day relative to expected
0.25/day are the major source of losses.
74
EFFECTIVENESS LOOP
. . . System continuously assesses State of
Control in manufacturing and Effectiveness of
control action taken.
i\share\sandys\conferences\informationengineering
-mit2.ppt
75
DATA TO DECISION CYCLE
DATA
INFORMATION
EFFECTIVENESS
MEANING
. . . did it work
MESSAGES
DECISION
ACTION
i\share\sandys\conferences\informationengineering
-mit2.ppt
76
. . . Value of Information is its Meaning
Content!
i\share\sandys\conferences\informationengineering
-mit2.ppt