requirements analysis and collection
TRANSCRIPT
-
8/14/2019 Requirements Analysis and Collection
1/32
(c) Kim H. Pries & Jon M. Quigley 1
Product requirementsCollect and analyze product demands
From the authors of:
http://search.barnesandnoble.com/Six-Sigma-for-the-Next-Millennium/Kim-H-Pries/e/9780873896566/?itm=1 -
8/14/2019 Requirements Analysis and Collection
2/32
(c) Kim H. Pries & Jon M. Quigley 2
Overview
What are requirements
Who are the stakeholders
Stakeholder management
How to gather requirements
Analyze requirements
Document requirements
Change Management
Verify requirements are met
-
8/14/2019 Requirements Analysis and Collection
3/32
Background
Trucks / automotive ($80K and up)
Used to make money
Numerous Electronic Control Units
Large variation in feature content and ECUs
(different applications)
High use / demand (130Kmi/year)
Extreme environmental exposure
-
8/14/2019 Requirements Analysis and Collection
4/32
(c) Kim H. Pries & Jon M. Quigley 4
What is a Requirement
A description of a feature or featuresexpected to provide a benefit
Benefit may not always be clear
The features the product must have to beaccepted by the customer
This is the scope of the product / project
The driving force behind the ROI for thecompany
-
8/14/2019 Requirements Analysis and Collection
5/32
(c) Kim H. Pries & Jon M. Quigley 5
Requirement types
Performance requirements
Design to requirements
Functional requirements
Derived requirements
Irrelevant requirements
Incorrect requirements
Redundant requirements
-
8/14/2019 Requirements Analysis and Collection
6/32
(c) Kim H. Pries & Jon M. Quigley 6
What is a good requirement
Cohesive
Complete
ConsistentCorrect
Observable
Feasible
Unambiguous
Required
Verifiable
Traceable
-
8/14/2019 Requirements Analysis and Collection
7/32
(c) Kim H. Pries & Jon M. Quigley 7
How to gather requirements
Direct customer interviews
Regulations or standards
Competitive assessments
Sell a great idea
Functional decomposition Organized and hierarchical
May serve as a requirement poka-yoke
-
8/14/2019 Requirements Analysis and Collection
8/32
Know the stakeholders
Internal / external
Supplier / Customer
Direct / Indirect Real / False
Continuous /
Intermittent
Severe / Cursory
Correlated / Inverse
Consider using anadaptation of theResource AllocationMatrix to manage
(c) Kim H. Pries & Jon M. Quigley
-
8/14/2019 Requirements Analysis and Collection
9/32
Stakeholder Management
Negotiationstrategies
Communicationstrategies
Leverage strategies(Political strategies)
Participatorystrategies
Compensationstrategies
Knowledge
Gathering/Sharingstrategies
Ignore-based strategies
Removal strategies
(c) Kim H. Pries & Jon M. Quigley
-
8/14/2019 Requirements Analysis and Collection
10/32
Stakeholder Management
(c) Kim H. Pries & Jon M. Quigley
RemovalIgnore-based
Knowledge Gathering/Sharing
Compensation
Participatory
Leverage (Political)
Communication
Negotiation
Internal / External
Supplier / Customer
Direct / Indirect
Real / False
Continuous / Intermittent
Severe / Cursory
-
8/14/2019 Requirements Analysis and Collection
11/32
Stakeholder Allocation MatrixCom
ponent2SW
Componen
t2HW
ProjectM
anager
Componen
t1SW
Componen
t1HW
SystemSpec
Systemver
ificatioTools
Manufac
turing
Aftermarket
T
raining
Tech-Pubs Component2 SW RD
Design Doc Component2 HW SR
Component2 A A In In In In In In Project Manager AC
Systems In In In In In A In In Component1 - SW MBComponent1 In In In A A In In In Component1 - HW SA
Systems Spec WC
Hardware A In C C C In C In In In System verification MMSoftware A In C C C In C In In In Tools JG
Manufacturing BW
Hardware C In A C In C In In In Aftermarket SRSoftware C In A C In C In In In Training BB
Testing Tech-Pubs LO
Component2 A A In In In
Component1 In In A A In
Systems In In In In ATraining
Component2 A A R R R
Component1 A A R R RSystems A R R R
Accountable A
Participant P
Review Required RInput required I
Sign off required S
Informed In
Consulted C
Design Doc
Component2 release
Component1 release
-
8/14/2019 Requirements Analysis and Collection
12/32
(c) Kim H. Pries & Jon M. Quigley 12
Requirements sources
RequirementsRegulatory
N
oisecontrol
Technology
con
straints
Marketexpectations
R
equirements
elicitation
Ethn
ographic
observation
-
8/14/2019 Requirements Analysis and Collection
13/32
(c) Kim H. Pries & Jon M. Quigley 13
Gathering Requirements
Stakeholder and Customer reviews
Focus groups
Field investigations
Simulation
Product mock ups (computer based instead of
embedded)
Rapid prototyping
Can document with a stakeholder matrix
-
8/14/2019 Requirements Analysis and Collection
14/32
(c) Kim H. Pries & Jon M. Quigley 14
System modeling
Aids creation of performance specifications
Allows for high-level simulations
Generate a working virtual mockup
Some of it can be done in spreadsheet
Provides holistic assessment of product performance
and subsystem interactions
-
8/14/2019 Requirements Analysis and Collection
15/32
(c) Kim H. Pries & Jon M. Quigley 15
System modeling limits
All can not be known in the early stages
If the product is novel this lack of knowledge is evengreater risk
-
8/14/2019 Requirements Analysis and Collection
16/32
Possible modeling techniques
DiagramsUse Case, class, etc.
Data flow
Flow chart
Pseudocode
ITU Z.100 SDL
Z, B-method, VDM, and other formallanguages
(c) Kim H. Pries & Jon M. Quigley 16
-
8/14/2019 Requirements Analysis and Collection
17/32
(c) Kim H. Pries & Jon M. Quigley 17
Analyze Requirements
Non-Requirement
Requirement
Evaluation
Requirements
Documented
Requirements
Gathering
-
8/14/2019 Requirements Analysis and Collection
18/32
(c) Kim H. Pries & Jon M. Quigley 18
Analyze requirements
Competing / conflicting stakeholder inputs
Prioritize requirements
Match concepts to prioritized requirements
-
8/14/2019 Requirements Analysis and Collection
19/32
(c) Kim H. Pries & Jon M. Quigley 19
Requirements / Constraints
Complexity issues
Hierarchical problems (architecture)
Large modules vs small modules
Assigning development durations/staffing
Technical issues (language, compiler, etc.)
-
8/14/2019 Requirements Analysis and Collection
20/32
(c) Kim H. Pries & Jon M. Quigley 20
Concept Evaluation
Pugh analysis
Quality, Function,Deployment
Requirements andconstraints matrix
Pugh MS Office
Excel 97-2003
http://www.ciri.org.nz/downloads/Requirements%20and%20Constraints.pdf
-
8/14/2019 Requirements Analysis and Collection
21/32
(c) Kim H. Pries & Jon M. Quigley 21
Document requirements
If broken down into multiple documents,need a document tree to manage it
Follow a standard format
Identify patterns if meaningful
Consistency
-
8/14/2019 Requirements Analysis and Collection
22/32
(c) Kim H. Pries & Jon M. Quigley 22
Document Requirement
Very specific language clearly written
Individually identified alpha-numericdesignator
REQ_SW_HMI_20
Can get complicated when requirement
meets multiple needs Disambiguate with disaggregation
-
8/14/2019 Requirements Analysis and Collection
23/32
(c) Kim H. Pries & Jon M. Quigley 23
Configuration Management
Requirements form basis of configurationmanagement
Requirements met in each Software package (SW release notes)
Hardware package (HW release notes)
Product configuration (CMP) Test configurations (test coverage)
Configuration audit
-
8/14/2019 Requirements Analysis and Collection
24/32
(c) Kim H. Pries & Jon M. Quigley 24
Requirements Management
Requirements
DocumentUpdated
AdditionalRequirement(scope)Identified
ChangeManagement
ExcludeRequirement
-
8/14/2019 Requirements Analysis and Collection
25/32
(c) Kim H. Pries & Jon M. Quigley 25
Change Management
Change happens! Scope impact
Well defined at project start
Process Areas of responsibility (authority)
Create formal configuration management plan
Requirements themselves should fall under configurationmanagement
All change to receive cost/time quotations
Agreed upon by all parties
-
8/14/2019 Requirements Analysis and Collection
26/32
(c) Kim H. Pries & Jon M. Quigley 26
Change management points
Ability to flex with change can be a competitive edge
Product launch process should provide support andcontrols for change
Contingency plans for obvious potential change points
Product simulation can provide early testing of ideas
-
8/14/2019 Requirements Analysis and Collection
27/32
(c) Kim H. Pries & Jon M. Quigley 27
Verify requirements
Difficult if tracing back to potentially nebulouscustomer needs
At least a one-one correspondence with customer
specification
Derived requirements will break one-one
-
8/14/2019 Requirements Analysis and Collection
28/32
(c) Kim H. Pries & Jon M. Quigley 28
Requirements and Verification
Basis for verification tests
Compliance testing used to verifyrequirements met
More sophisticated approaches needed tovalidate Testing to customer needs
Testing to failure and, maybe, destruction
-
8/14/2019 Requirements Analysis and Collection
29/32
(c) Kim H. Pries & Jon M. Quigley 29
Traceability
Quality Function Deployment supportstraceability at a high level
Can also build traceability matrices
Surest way to confirm requirement are met
A solution for each requirement
Complete
Traceability matrix indicates requirements met butgive no idea of optimality
-
8/14/2019 Requirements Analysis and Collection
30/32
(c) Kim H. Pries & Jon M. Quigley 30
Requirement Verification
Test cases for each individual Requirement TC_REQ_SW_HMI_20:1
TC_REQ_SW_HMI_20:2
Confirms product delivered meets the
customer demands
-
8/14/2019 Requirements Analysis and Collection
31/32
(c) Kim H. Pries & Jon M. Quigley 31
Feedback
Compare requirement to delivery
Use initial requirements list to control scopecreep
Consider expressing requirements in aformal language (VDM, Z, structured
English)
-
8/14/2019 Requirements Analysis and Collection
32/32
(c) Kim H. Pries & Jon M. Quigley 32
Additional Information
Jon Quig ley:jonmquig [email protected] om
Kim Pries: [email protected]
mailto:[email protected]:[email protected]:[email protected]:[email protected]