What made me a Systems
Engineer…
Dr Kevin Howard
Copyright © 2014 Dr Kevin Howard.
Published and used by the SSSE and INCOSE with permission. The author or assignee retains the copyright to the materials.
A personal perspective
• No excuses – this is personal
• Its always good to learn from mistakes • Particularly other peoples!
• It can be useful and interesting to look back at experiences
The plan
• Quick review of background
• The things have contributed to my systems understanding
• Pick on some ‘learning experiences’
• Thinking failures
• Process failures
My Background
• Radars:
• High-power amplifiers, receivers, waveform design, imaging
• Antennas:
• Printed, conformal, arrays, large scan angle
• Software:
• Telephony, data fusion, DSP, spread-spectrum
• mmW
• Communications, passive imaging, direction finding
• Communications
• DBS, subterranean, spread spectrum, covert
• Vehicles
• Chief SE Foxhound AFV, Chief Engineer Warrior AFV ... Maintaining 40+ year old cars
• SE Consultancy
A systems perspective on learning
• Stakeholder management
• Requirements management
• Technical development
• Integration
• Testing and evaluation
• Process
• Organisation
Stakeholder management
The requirement
• Red Phone – early ‘90s
• Customer: ITV (UK)
• Requirement – 100 way emergency teleconference in 3 seconds
• … In case the Queen dies!
…In case the advert machine breaks
• Solution ICT: 22x 486 PC, (up to 5 channels each site), T1 ISDN dial-up, PCI conferencing card,
• Customer
• A lose collection of franchises trying to do a single procurement
• Open competition
The Lesson…
The lesson
• Not loud enough, too loud, not intelligible, not red enough
• Requirements are not enough
• Understand and manage your customer
• People like to feel engaged
And it all comes out at acceptance…
Requirements capture
• ESM system with coverage
• Antenna 4 Steradian coverage
The Lesson...
To Mis-quote Scotty…
• Ye cannae change the laws of physics
Systems Engineers are people
not experts
Technical Development
The requirement
• Integrate next generation data fusion with 1970s radar
• 10MHz system clock
• 18 bit data
• Connecting to Pentium
• 300MHz clock, 66MHz FSB
• 32bit
• PCI I/O card clocking at 100MHz
The Lesson... Never underestimate technology
• 20way multi-coax connection
• Unfamiliar so they get damaged easily
• Can’t see the shorts went the cables get crushed!
• Technology turns over quickly
• People forget
• Failure to assess data throughput of wire wrap technology
• Slow clock... But massively parallel
• 17 synchronous cabinets
• No OS overhead
Sub-system dependencies
The requirement
• Very wide band amplitude Direction Finding (DF)
• Orthogonal spiral antennas
• 10:1 bandwidth
• Self similar pattern
The Lesson…
Technology is never perfect
• Above 3:1 Bandwidth higher order modes interfere
• Beam is not circular
• It rotates with frequency
… the tracking law varies with frequency
• Correction demands greater precision in frequency discriminator
… Expensive
• A new spiral geometry
Integration, Test and Acceptance
Back to Red Phone
• Getting 100 people to answer the phone in 3 seconds is a challenge
• A quick, low cost answer
The Lesson
• 100 channels of coherent source
• Very loud speaking clock
• Can miss lead the test process
• Digital exchange interprets rapid dialing of the same digits as panic action
…And automatically calls the police
Process
• Providing service feed to a SA system:
• Clear spec, clear objective, clear understanding ... Should be straight forward
• Process context
• Subcontractor operating at CMMI level 3
• Prime at operating CMMI level 5
Lesson
• Little-Endian data feed instead of specified Big-Endian
• Single bit error in thousands of lines of code
• Discovered at integration testing
• Prime asks sub to change spec and code ‘because it’ll take you an afternoon. It’ll take us 2 months’
Don’t think process will fix everything
Organisation
• System Engineering managed under 29(!) working groups (WG)
• Poor communication between them
• Dysfunctional engineering operations
• Programme threatened by lack of progress ...Some were just talking shops
The lesson
• Deleted and merged WGs until only 8 left
• Forced engineers to communicate with each other!
• Forced issues to be resolved
• Ensured clear direction and focus
Talking is not communicating
What about System of Systems Engineering
• The planned combination of assets to achieve an objective or
purpose where the assets were NOT contrived for the purpose were not necessarily design to work together.
• Operate a capability level
• Its about interfaces
• And information
• But mainly about
What have I learned...?
Requirements Management
• Every Requirement adds cost
• They don’t all add value
• Delete requirements where you can
• Death, Taxes and Requirements change
• Its the way of the world .... It’s not a risk
• Plan and budget for change
Not just change management Companies make a living on it!
OGIDTOT
Oh God I Didn’t Think of That
However well you plan there will always be
something you didn’t expect
Its called risk and its unavoidable
‘... But don’t put any risks in that will affect the
programme...’
What is Systems Engineering?
‘Systems Engineering is an interdisciplinary approach and means to enable the realization of successful systems. It focuses on defining customer
needs and required functionality early in the development cycle, documenting requirements, then proceeding with design synthesis and
system validation while considering the complete problem’
The combined interaction of
technologies to achieve a specific objective
How is Systems Engineering Achieved?
‘Systems Engineers are there to sustain an environment of communication between disciplines and stakeholders that fosters joint working and understanding’
Its about people…
…We employ them for what they know
Its about language
• Whether that’s written or graphical
Class: Amplifier Gain: BW: Class: Circulator Class: Amplifier Gain: BW: Class: Antenna
