Major Challenge in Systems Engineering
• One of the major challenges in
engineering systems is to get all of the
parts, components and subsystems to
work together
• Getting them to work together and
understanding/controlling their interaction
is called Systems Integration
1
Major Obstacle To System Integration
• One of the major obstacles to successful system integration is the
fact that a big part of the job in building systems is done below the
subsystem level
• This is called “discipline” engineering because it often utilizes
specific disciplines
– Power engineering
– Structural engineering
– Mechanisms engineering
– Data Systems Engineering
– Etc…
• Because most engineers are trained in disciplines, they tend to pick
up that discipline as their world view
– “When your only tool is a hammer, all of your problems begin to look like
nails”
• Systems engineering is the discipline which enable us to look at the
whole system and understand its integration and interactions
2
3
When one requirement dominates
design, you rarely get a world champion
design
4
Supermarine Spitfire
Mission – Fighter Aircraft
Optimized for aerodynamic
performance
Suboptimal – stability – nasty
spin mode
GeeBee
Mission – Racer
Optimized for engine and
minimal drag
Suboptimal - controllability
5
YB49 Flying Wing (1940’s)
precursor to modern B-2
Mission - Long Range Bomber
Optimized for range by
aerodynamic performance and
large wing tanks
Suboptimized for stability and
control
Gossamer Albatross
Mission – Human Powered Flight
Demonstrator
Optimized – Minimum Weight and
maximum lift so that vehicle could
fly within performance of human
engine
Subotimized – strength – very
fragile
6
Phenomena of discipline view
happens in all systems
7
Important Realization For A Systems Engineer
• There is no one single best design for an airplane,
spacecraft, missile or anything else
• The measure of “goodness” is how well it meets a
specific requirement
– An airplane that is a great fighter makes a lousy
cargo carrier and vice versa
8
Evolution of Systems Engineering
• One of the primary sources of systems engineering was American
Telephone and Telegraph (AT&T)
– Research arm of AT&T was Bell Telephone Laboratories (“Bell Labs”)
– Western Electric manufactured everything for AT&T, including
prototype hardware
– Bell Labs developed a method of formal specifications for
communicating about prototype hardware
• MIT was responsible for developing radar systems and worked with Bell
Labs on gun fire control systems during WWII
• MIT had significant problems integrating the MIT radar and Bell Labs gun
control system on a project called the SCR-584 Fire Control System
– They soon realized that the electronics acted differently when
integrated together then when they were tested separately
• As these systems evolved during the war, it became clear that a systems
integrator was required
9
Western Electric Telephone – ca 1949
Mk 56 Gun Fire Control System
10
Evolution of Systems Engineering
• MIT (Ivan Getting) took this role on in 1945 on the Mark 56 gun control
system
– Adopted Bell Labs specification techniques
– Became the center of a web of communications between many
contractors and the users
– Established a formal system of systems integration
• Getting took these ideas to the USAF through his role on the USAF Scientific
Advisory Board and in the courses he taught at MIT
• These ideas were incorporated by Simon Ramo in the AIM 4 Air to Air Falcon
Missile
• Ramo-Woolridge (later Thompson-Ramo-Woolridge- TRW) brought these
ideas to USAF ballistic missile development in the 50’s and 60’s
• These ideas then moved into the US Manned Space Program and went
everywhere from there
11
12
Evolution of Systems Engineering
AIM4 Falcon Missile
Air To Air Missile
Mercury/Atlas,
Gemini/Titan and
Apollo/Saturn
Atlas ICBM Titan ICBM Minuteman 13
ICBM
Systems Integration –
The Vee Model
• The Vee Model addresses the technical aspect of the project cycle
and represents the sequence of project events.
• The VEE model starts by determining the high level systems
requirements. A system design at the high level is established which
allocates functions and requirements to subsystems.
• Those allocated system requirements are broken down into specific
requirements for subsystems, components and parts.
• Specific designs are done for the subsystems, components and
parts and then the parts are manufactured and assembled.
• Parts and components are tested to see if they meet the component
and part requirements.
• Parts and components are assembled into subsystems which are
tested to subsystem requirements..
• Then the entire system is assembled and tested to the top level
system requirements.
14
A CI is a “Configurable Item” – basically a product that is 15
delivered as a complete unit. A system is made of many CI’s.
What’s Wrong with the VEE ?
• All the integration is pushed off to the end
– Integration is usually the most difficult part
– Hard to maintain support for a project when all
you can demonstrate is pieces
– Hard to iterate with users/get feedback on pieces
• The need to integrate early forced
evolutionary development approaches
– Intermediate levels of development which are
integrated and tested
• Fully developed, this led to rapid spiral
development methodology
16
What’s Wrong with the VEE ?
http://libraries.maine.edu/Spatial/gisweb/spatdb/urisa/ur94007.html 17
Rapid Spiral Development Methodology
• Mature system through building, testing and using a series of
prototypes of increasing complexity until all of the required
functionality is achieved
– Originally developed by JPL in late 80’s early 90’s
– Used by NASA in the new Mission Control center and then in
X-38 project
– Parallel development in DOD led to baselining spiral
development methodology in major programs
• Tends to work best in software projects with short lead times
and small physical assets
– Can work well in hardware development projects – X38
– Have to be able to build things with sufficient functionality to
integrate and test
18
Rapid Spiral Development - Wright
Brothers
Overview of Wright Brothers Discoveries
http://wright.nasa.gov/discoveries.htm 19
20
21
Any Questions ?
22
