Quality in Architecture-Centric Engineering

Quality in Architecture-Centric Engineering
presentation to INCOSE HR Area Chapter
By
Christopher P. ‘Jake’ Jacobson
23 February 2011
1
Overview
• Background
• Scenarios and Performance Measures
• Applied Solutions and Lessons-Learned
2
Background
•
•
•
•
•
NORAD Space Surveillance – HIPO
USAF Air Surveillance – A, B Specs
Joint Staff GCCS – Enterprise Architecture
USS Cole/9-11 – Capability Spirals
Enterprise Service Integration &
Interoperability Lab (ESIIL) – Integrated
Tools & SoSE
• Systems & Software Technology
Conference (SSTC) ’08 - ACE
3
Significance of ACE
4
Building Blocks
Artifact Examples
Class
Collaboration
Diagrams
Diagrams
Use Case
Diagrams
Data
Models
Data Dictionary
Activity
Models
EventTrace
Diagrams
State
Chart
Colored
Petri
Net
Process
Flow
Simulator
State
Machine
• Different bands enable
improved technique as
alternative to natural
language
• Transition between levels
of abstraction helps
transform requirements to
design
• Behavior models provide
benchmark for
comparative analyses
based on measures
5
Adapted Engineering ‘V’
• Downstroke creates
Fully Discovered ‘asis’ and ‘to-be’
Baselines
• Upstroke realizes
Enterprise Solution
• Apex defines
Engineering Choices
against Performance
Measures
6
Scenarios & Performance Measures
Static
Breadth
Systems
EventHandling
Systems
GIS
Systems
eXPanel
Information
Decision
Support
Systems
Orion
WGC
GeoSpatial
Depth
UICDS
ESRI
IBM
WBM
Plans
Unit
Plans
People
State
EOC
Reports
Event
Type
Checklists
Shelter
Plans
District
Red
Cross
Staging
Report
Shelter
Lead
FEMA
Region
III
County
EOC
County
Planner
Dynamic
To Behavioral
Models &
Implementation
Engineering
7
* Support to NRF protection, preparedness, response and recovery operations
Architecture, Requirements and
Performance Measures
• Requirement:
Threshold: Handle 10
support requests per hour
• Executable Architecture
& Workflow:
Objective: Handle 20
support requests per hour
QUID Identifier digitally
connects requirement to
Event Trace(s)
8
20
Derived Measurements
Quantitative Probe:
Objective Run – 20 Requests/Hour;
QUID Links Scenario Run to Requirement
• Max resource for min
bottleneck baseline run
Sustainable
Baseline
Performance
@ 5 Requests/Hour
Support Request Performance
120
5 Requests/Hour
• Resources used; set
resource constraint
• Increase token rate for
objective & threshold
• Fixed resource for
sustainable increased
rate alternatives
100
Time (Minutes)
10 Requests/Hour
80
20 Requests/Hour
60
Required, Baseline &
Alternative Performance
Linear (5
Requests/Hour)
40
Linear (10
Requests/Hour)
20
Linear (20
Requests/Hour)
0
1
2
3
4
5
6
# Jobs
7
8
9
10
Alt 1: Alt 2:
• TOC
• Init $
• Risk
• ROI
• TOC
• Init $
• Risk
• ROI
RATE (minutes)
9
21
Applied Solutions & Lessons Learned
Air & Space Operations Center (AOC) WS – Targeting Improvement Spiral
Staffing (TDO/Ts)
TBONE DMPI Processing
6.25 hr Allocation against 12 hr Plan/Exec
(5 DMPIs/Target)
60
50
40
30
20
10
0
TBONE As-Is
TBONE Gridlock
760
1500
2435
DMPIs
Joint Forces Command (JFCOM) - Joint Close Air Support (JCAS) Study
Free Float Time = FOUO
Net-Enabled C2 (NECC) – Air Support Analysis
10
17
Customer Acceptance &
Repeatable Scenario-Driven Approach
11
Quality Improvement &
Techniques for Quantitative Measurement
12
Experience & Lessons Learned
• AOC WSI – Dynamic Targeting Spiral Decisions;
– Linkage between SoSE and Product Integration teams
• JFCOM/J89 EA – JCAS Desk-Top Analysis (DTA)
– Associate Contract Agreements and Shared Portal
• DISA NECC – ISP (ASR Vignette) JS/OSD Approval
– Shared tools environment for Engineering Risks/KPIs
13
Additional Information
• Chris Jacobson – chrjacob@us.ibm.com
• Achieving Interoperability for Critical IT
and Communication Systems, Artech
House Publishers
– http://www.amazon.com/AchievingInteroperability-Critical-CommunicationSystems/dp/1596933895
14