“An Empirical Investigation of System Changes to Frame Links between Design Decisions and Ilities” J. Clark Beesemyer, Adam M. Ross, and Donna H. Rhodes Massachusetts Institute of Technology CSER 2012 St. Louis, 19-22 March 2012 Motivation “Rep. John Mica called on the agency to "reform" and "become...a thinking, risk‐based, flexible agency that analyzes risks, sets security standards and audits security performance.” “Defense Secretary Panetta: "The US joint force will be smaller and it will be leaner. But it will be more agile, more flexible, ready to deploy quickly, innovative and technologically advanced.“ … “the Defense Department and the Office of the Director of National Intelligence pledged to foster an industrial base that is 'robust, competitive, flexible, healthy, and delivers reliable space capabilities on time and on budget.'" Quotes from AIAA Daily Launch, 20 Jul 2011 – 13 Feb 2012 Ross, Beesemyer, and Rhodes 2012 seari.mit.edu © 2012 Massachusetts Institute of Technology 2 Motivation • High impact early decisions usually made with incomplete system knowledge • Decision makers may improve their capacity to discriminate between system concepts and design choices by measuring a system’s “ilities” such as changeability, scalability, and survivability – These system properties (ilities) can be ambiguous, imprecise, and hard to validate or verify • Increased interest in ilities requires better understanding of the change mechanisms and more precise language Looking at empirical cases of system changes and associated ilities in a structured manner may lead to more effective design approaches. seari.mit.edu © 2012 Massachusetts Institute of Technology 3 Ilities as Outcomes Design Principle Path Enabler Change Mechanism Ility Characteristic • Ilities are performance characteristics made possible ultimately by the design principles used in the early phases of design • Descriptive research aims to look at many of these cases for trends and insight that may be used in future designs Targeted Modularity seari.mit.edu Software packaged add-ons (apps) Downloading from AppStore © 2012 Massachusetts Institute of Technology Flexibility Versatility Extensibility Evolvability 4 The Change Option WP-2011-1-2 Ross, A.M., and Rhodes, D.H., “Anatomy of a Change Option: Mechanisms and Enablers”, <seari.mit.edu/papers.php> S2 S1 $.initial_cost M1 S3 S4 Path Enabler Restrictive Epoch (num_uses>1) S2 M1 M1 S1 S1 M1 M1 M1 $.carry_cost $.execution_cost Start date Expiry date Time The following factors characterize a change option: • • • • • • Name Start date Expiry date Possible end states = f(start state) Initial cost = cost of path enabler? Carry cost = f(now, execution date, expiry date) • • • • • • Execution cost = f(end state, epoch) Reusability = number of times it can be executed Valid epochs Valid lifecycle phase Pre-requisites for execution (e.g. agent) Change “type” (e.g. flexible, adaptable, etc.) Change options can be compared and valuated on this basis seari.mit.edu © 2012 Massachusetts Institute of Technology 5 Generalizing the Change Statement: A Prescriptive Basis (From Ross, Rhodes, and Hastings 2008) 1 2 3 4 5 6 7 8 9 10 10 dimensional basis for specifying “changeability”-related ilities WP-2011-1-2 Ross, A., Beesemyer, J., and Rhodes, D., “A Prescriptive Semantic Basis for System Lifecycle Properties”, seari.mit.edu/papers.php seari.mit.edu © 2012 Massachusetts Institute of Technology 6 Generalizing the Change Statement: A Prescriptive Basis 10 dimensional basis for specifying “changeability”-related ilities 1 2 Dim # 1 2 3 4 5 6 7 8 9 10 3 4 Category Name Cause Context Entity Aspect Phase Agent Parameter Type Effect Potential States Valuable 5 6 7 8 9 10 # Levels 4 3 4 4 4 4 3 5 4 256 WP-2011-1-2 Ross, A., Beesemyer, J., and Rhodes, D., “A Prescriptive Semantic Basis for System Lifecycle Properties”, seari.mit.edu/papers.php seari.mit.edu © 2012 Massachusetts Institute of Technology 7 Captured Ilities Prescriptive Semantic Basis for Changeability-type Ilities In response to “cause” in “context”, desire “agent” to make some “change” in “system” that is “valuable” Cause Context System Agent Change Valuable Why Where What What When Who What What What What When When For What For What Cause Context Entity Aspect Phase Agent Param Change Type Effect (Scale) Effect (Amount) Potential States Timing Span Resources Benefit perturbation specificity abstraction aspect LC phase executes param type level set target range reaction duration cost utility disturbance shift none any circumstantial general any architecture design system any form function operations any pre-ops ops inter-LC any internal external either none any level set any bigger smaller not-same same any more less not-same same any one few many any sooner later always any shorter longer same any less more same any more less same any (choose one) (choose one) any any shift shift shift any any any any any disturbance any any any any any shift any any circumstantial circumstantial general any any any any any circumstantial any any any any any any seari.mit.edu any any system system architecture any any any any any any any any any any any any function operations any form any any any any form any any form any any any any any ops ops ops ops inter-LC any any any ops any ops ops any any ops any any any any any none any internal external any any any any any any any any either any set set any level any any any level set set any any any any set any any any any same same any not-same not-same not-same any any any any not-same not-same any not-same any not-same not-same same any any not-same not-same any not-same not-same any any not-same not-same more not-same any Ility Name many many any few any any any any many any any any any any any any any © 2012 Massachusetts Institute of Technology any any any any any any any any any any any any any sooner any any any any any any any any any any any any any any any shorter any any any any any any any any any any any any any any any any any any any any any any any any any any any any any any any any any any any any any same Functional Versatility Operational Versatility Robustness Classical Passive Robustness Evolvability Adaptability Flexibility Scalability Optionability Modifiability Survivability Reconfigurability Agility Reactivity Extensibility Changeability Value Robustness 8 Change Database • To capture changes from many different systems and various domains – Currently holding ~100 changes from ~45 systems • Developed from Ility Framework previously discussed • Takes system and change information in a categorical manner to determine ilities present in changes seari.mit.edu © 2012 Massachusetts Institute of Technology 9 Change Database This research aims to find a better means of determining which ilities are present in different system changes and map those ilities to various design principles. When stakeholders identify an ility as a desired property of a design, the ultimate goal of added value to the system from having this ility represents one end of this relationship. seari.mit.edu © 2012 Massachusetts Institute of Technology 10 Change Database Path Enablers are instantiations of DPs that facilitate CEs in a system. Design Principles are the heuristics designers follow when making design choices. seari.mit.edu www.militaryfactory.com Ilities are properties shown during these changes that equate to more or less value for stakeholders. Change Mechanisms give the system options to change if necessary given conditions and performance © 2012 Massachusetts Institute of Technology 11 Change Database Change database seari.mit.edu © 2012 Massachusetts Institute of Technology XXXXXXXXXX 12 Change Database System Information XXXXXXXXXX Source and Cost Info (if available) seari.mit.edu © 2012 Massachusetts Institute of Technology 13 Change Database Specified Parameter XXXXXXXXXX Path Enablers Ilities Identified seari.mit.edu © 2012 Massachusetts Institute of Technology 14 Change Database XXXXXXXXXX Change Information seari.mit.edu © 2012 Massachusetts Institute of Technology 15 Cluster Analysis COOLCAT • Clustering algorithm made to handle categorical data, (Fu et al.) • Model based clustering (entropy model) – Where entropy is a measure of dissimilarity • Basic Method – Turn categorical data into nominal dataset – Specify number of clusters – Run piecewise entropy to initialize first 2 clusters w/ highest entropy (most different) – Max(min(entropy)) of remaining records for each new cluster to specified number of clusters – Fill remaining records into clusters by minimizing system entropy for each record seari.mit.edu © 2012 Massachusetts Institute of Technology 16 Preliminary Insights Heuristic-based clusters COOLCAT clusters Cluster 1; Size = 28 Entropy = 5.616; System Entropy = 5.88 Finite 4 Cause Shift19 None 1 Shift28 Cause Context Circumstantial 5 General 10 Circumstantial 14 Context Entity Architecture 7 Design 4 System External Level 5 Set23 Agent External 18 20 4 Internal Finite Cause 0.4 None Shift Cause 0 0.1 0.2 0.4 0.3 Circumstantial 47 Finite 24 0.5 0.7 0.6 0.9 0.8 None py y Agent External 52 System11 Design5 0 0.1 Level 0.3 0.4 0.5 0.6 0.8 0.9 13 Phase E (ops) Inter-LC 0.2 Param_Type Level 9 14 Set 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.2 Many12 0.4 0.6 One 5 0.8 Operations2 Many5 0.4 0.6 None 1 Form 20 Aspect 0.8 0 0.2 0.4 0 0.2 Set Many18 0.4 0.6 0.8 0.6 0.8 1 Shift19 Form 16 2 5 Function Operations E (ops)23 External 22 Agent Level 5 0 Internal 1 Set18 Few 5 Potential_States 1 General 11 System 20 Aspect Param_Type One 10 0.8 Level Few 7 Potential_States 1 Set1 Circumstantial 12 Phase Set5 0.6 0.4 EntityDesign3 Internal 10 Many14 0.2 Cause Finite 4 Context E (ops)24 Potential_States 1 Shift11 Level 19 Param_Type 0.8 17 Cluster 6; Size = 23 Entropy = 5.384; System Entropy = 4.003 Operations4 External 14 Agent Few 4 0 1 One 6 0.6 8 Few 4 Potential_States 0.8 System 24 Phase Set5 0.6 0.4 Inter-LC8 External 25 Param_Type Set 0.2 2 3 Function Operations E (ops)17 Agent None 2 Level 3 Param_Type None 2 System 14 Form 20 Aspect Phase Many5 Few 0 General 13 Entity Architecture 2 Design9 11 Level E (ops)4 Internal 2 Circumstantial 12 Context Many17 0.2 0.4 0.6 One 1 0.8 1 Military vs non-military Control the number of clusters (2-n) Large population vs small population Mass produced vs unique Can compare mental clusters to COOLCAT clusters. Long lifespan vs short COOLCAT will find similarities more complex than simple data mining. For example, the mental models show that shorter lifecycles tend to be more evolvable. COOLCAT models yield an evolvable cluster that is almost equal lifecycles but much more military and space oriented. seari.mit.edu 1 Shift25 Cause General 1 Internal 5 Operations3 4 tential_States Agent Circumstantial 24 Context Entity Level 4 0.2 0.4 Finite 12 Cause General 5 External 9 Param_Type 0.2 Shift2 Operations9 External 8 System 3 4 Form 1 Aspect Cluster 5; Size = 24 Entropy = 4.555; System Entropy = 4.003 Inter-LC9 Phase One 1 Many11 Potential_States 0 0 1 Design Param_Type Phase Few 17 Potential_States 1 Design6 Form 7 Agent Few 11 0.8 Shift9 Circumstantial 4 Entity Architecture 3 Aspect Potential_States 0.6 Operations1 Internal 6 Level 17 Param_Type Many1 One 3 0.4 Cause Context External23 Agent Few 0 19 Cluster 4; Size = 25 Entropy = 5.87; System Entropy = 4.604 E (ops)15 Shift2 Agent Circumstantial 4 Entity E (ops)17 External 11 Phase Finite 2 Cause Context Form 16 Aspect Cluster 4; Size = 9 Entropy = 4.656; System Entropy = 4.003 10 Internal 14 Level 36 Potential_States 1 System 14 6 Aspect Cluster 3; Size = 4Form Entropy = 4.434; System Entropy = 4.003 Shift11 System 17 Agent Few 10 0 One 0.8 Circumstantial 14 Context Circumstantial 17 Phase Set14 Param_Type Operations6 Operations2 Inter-LC11 External 14 Agent 1 0.6 Finite 13 Cause 1 Finite 6 Entity Form 12 Phase E (ops)3 Potential_States Form17 Aspect 0.8 Context Architecture 14 Entity One 12 0.7 Many 0.4 3 Cluster 3; Size = 15 Entropy = 6.046; System Entropy = 4.604 One 11 0.6 Cause General 14 Aspect 0.2 None 2 Cluster 2; Size = 17 Entropy = 2.196; System Entropy = 4.003 Shift14 Cause Set Many31 2 Entity Design1 33 Few 24 Few External 22 Agent Param_Type 10 0.2 Set20 Many27 0.4 Cluster 1; Size = 14 Entropy = 2.436; System Entropy = 4.003 General9 tential_States E (ops)36 Phase Set15 0 Level 43 0.2 Context Param_Type Architecture 7 General 7 Inter-LC13 34 Inter-LC12 External 15 1 13 Function Operations Internal 19 Few 25 0 Internal 15 Operations2 E (ops)63 ntial_States 2 13 Function Operations Shift External 42 Param_Type E (ops)54 ential_States Entity System Phase Circumstantial14 Context E (ops)3 Agent System 62 Agent Form 52 General 1 System 36 Form 34 Aspect 1 Param_Type 56 Form 49 Phase py Shift21 0.8 38 Circumstantial Aspect 51 1 Aspect Cause Finite 2 0.6 1 Entity Design1 General 20 6 Many 10 6 One 2 EntityArchitecture Design Few 16 Potential_States Set10 Context Shift24 Circumstantial 35 Entity Operations3 Form 12 Aspect Phase 47 Context Level 14 Param_Type One 3 Cluster 2; Size = 63 Entropy = 5.997; System Entropy = 5.88 2 None 1 Context Architecture 15 Entity Many10 0.2 Inter-LC7 Finite 11 Cause General 15 Context 28 Few 15 ntial_States 0 E (ops)17 Cluster 2; Size = 36 Entropy = 3.923; System Entropy = 4.604 Shift15 Cause Inter-LC20 E (ops) Agent Operations4 Cluster 1; Size = 15 Entropy = 2.686; System Entropy = 4.604 System 2 1 4 Function Operations 8 Param_Type Phase Design9 Form 23 Phase 13 Form 20 Aspect General 23 Architecture 17 Entity Aspect © 2012 Massachusetts Institute of Technology 17 1 Significance • 2 clusters • 6 clusters – Expected system entropy 5.8508 using COOLCAT – Expected system entropy 3.6795 using COOLCAT – Exp Sys Ent 1000 trials using random clustering: – Exp Sys Ent 1000 trials using random clustering: 80 120 70 100 60 80 50 40 # of trials # of trials 60 40 30 20 20 10 0 7.3 7.4 7.5 7.6 7.7 7.8 7.9 0 6.7 6.8 seari.mit.edu 6.9 7 7.1 7.2 7.3 7.4 7.5 7.6 Expected system entropy Expected system entropy © 2012 Massachusetts Institute of Technology 18 7.7 Conclusion • Systems can use similar path enablers to achieve many different states, and these changes may be characterized by multiple ilities • Despite desire to implement ilities in designs, a clear method is still lacking for identifying “good” designs with respect to preferred ilities • Possible to classify / decompose system changes in a 10dimensional semantic basis to describe the change mechanism • Ilities are attributed not a priori, but as an outcome to the change • Enables decision makers to create better requirements that are clear, concise, and verifiable. Clarity in language is important as we look for design insights in groups of change mechanisms present in many systems. seari.mit.edu © 2012 Massachusetts Institute of Technology 19 Questions? seari.mit.edu © 2012 Massachusetts Institute of Technology 20
© Copyright 2026 Paperzz