Enterprise Opportunity and Risk B. E. White The MITRE Corporation 11 July 2006 Public Release Case Number - 05-1262
Relative Importance of Opportunity Uncertainty Risk Un-assessable Unknown Opportunity Enterprise Scale The minimum goal of this talk is to System of Systems Scale raise your sensitivity level for proactively pursuing opportunities at all engineering scales. Systems Scale Risk Opportunity 2 See Notes Page
Risk/Opportunity Representation on Probability/Impact Grid “Attention Arrow” High High Probability Probability P o Q o Medium Medium Risks Opportunities Low Low Low Medium High High Medium Low Positive Impact Negative Impact Benefit of Success Consequence of Failure B s C f 3 ___________ * After [Hillson, 2004], p. 126 See Notes Page
What Are Consequences of Failure?* THEN these are the consequences Condition Root Cause Consequence Present 1 Event 111 CONDITION Subsystem will reveal unanticipated Current test plans are focused on the components of performance shortfalls Event B the subsystem and not on the subsystem as a whole. IF this Risk Consequence Event 211 Event A Occurs Full-up system will reveal Risk Event 11 unanticipated performance shortfalls Subsystem may not be fully tested when integrated into the system Consequence for full-up system-level testing. Event 311 Subsystem will have to incorporate late fixes to tested software baseline The region bounded by this Consequence Event 411 space is Probability (A|B) Subsystem will have to accommodate unanticipated changes in subsequent build Consequences of failure are undesirable hardware/software requirements which will events that degrade the performance or affect development cost and schedules capability of a system, SoS, or Enterprise. Consequence Event 511 The Risk Statement: User will not accept delivery of subsystem hardware/software without fixes An Illustration of CONDITION-IF-THEN 4 ___________ * [Garvey, 2005], p. 7 See Notes Page
What is Opportunity? • Opportunities are events or occurrences that assist a program in achieving its cost, schedule, or technical performance objectives. • In the larger sense, explored opportunities can enhance or accomplish the entire mission. • Opportunity also is associated with uncertainty and impact. • There is a duality or parallelism to risk that can be applied. • For an opportunity, let Q o be the probability of occurrence, B s , the benefit of success, and E e , estimated enhancement. • We can pose the simple formula: E e = Q o × B s [This is expected benefit!] 0 < B s < ∞ Opportunity Assessment Benefit = A o = {Q o , B s } An interpretation: No Gain Worthwhile Pain Golden Opportunity Windfall Euphoria Probability = 0 < Q o < 1 5 See Notes Page
Opportunity Classification Example* Opportunity “Opportunity Averse” System Profile B s “Opportunity Seeking” System Profile Opportunity Q o Opportunity 6 ___________ * After [Garvey, 2005], p. 8 See Notes Page
Thoughts About Opportunity and Risk Concerning TSE, SoS Engineering, and ESE or CSE • Think about opportunity/risk with respect to a complex system’s environment in addition to the system per se . – There may be many more opportunities in the system’s environment. – The pursuit of these opportunities could reduce the system’s “stress”. – Environmental risks seem less important than the opportunities. – Enterprise-scale opportunity action and risk avoidance can be viewed with a philosophy of “nothing ventured, nothing gained”. – Downside risk is about not incurring “damage” that might stifle the aforementioned opportunities. • Compare and contrast TSE and CSE concepts. – A complex system (and enterprise) is “open”. – This suggests a predisposition for opportunities. – One should “open” the system further to create more emergent behavior. – Be more aggressive with identifying, exploring, and developing opportunities than in TSE. 7 See Notes Page
Thoughts About Opportunity and Risk Concerning TSE, SoS Engineering, and ESE or CSE (Concluded) • Enterprise risks can be mitigated by creating a management process that has built-in abilities to – Quickly assess whether emergent behavior is desirable – Encourage desirable behavior – Discourage undesirable behavior – Encourage greater acceptance of risks • Stevens: Messy frontier – Political engineering (power, control…) – High risk, potentially high reward – Foster cooperative behavior • One may learn from researching what economists do about opportunity and risk at multi-scales of analysis, i.e., macroeconomics and microeconomics.* • In summary – Opportunities for intervening in enterprise environments are great. – The greatest enterprise risk may be in allowing this process to atrophy. 8 _________ See Notes Page * [Kuras, 2004]
Regimen for Complex Systems Engineering (CSE) Analyze and Shape the Environment Tailor Developmental Methods to Specific Regimes and Scales Identify or Define Targeted Outcome Spaces Establish Rewards (and Penalties)* Judge Actual Results and Allocate Rewards Formulate and Apply Developmental Stimulants Characterize Continuously** Formulate and Enforce Fitness Regulations (Policing) _______________ 9 * Relates explicitly to CSE Opportunities and Risks See Notes Page ** Relates explicitly to CSE Opportunities
Opportunities and Risks in “Establish Rewards” • Suppose a suitable outcome space has been identified. • Autonomous agents will develop specific outcomes taking advantage of opportunities. • There is risk in developing products that – May not become outcomes – Become less desirable outcomes • These risks are either not rewarded or are rewarded less. • Because a reward is granted to many outcomes, agents may pursue opportunities more aggressively than mitigating the risks of not achieving outcomes. • Risk mitigation could be reduced to ordering outcomes according to rewards. • This ordering might be pursued in conjunction with other autonomous agents because rewards are granted only to targeted populations of agents. • The hypothesis that opportunities would be treated more aggressively than risks still needs validation. 10 See Notes Page
Opportunities in “Characterize Continuously” • This CSE activity is the continual generation and refinement of complex-system characterizations. Continuous Characterization is crucial for autonomous agents to independently develop metrics to guide their local decision making to be congruent. • The specific outcomes used as the basis for Judging should be characterized, as should the rationale that eventually explains the subsequent Judging decisions. • Rewards (and perhaps Outcome Spaces) initially should be characterized with succinct “bumper - sticker” labels. The U.S. Army motivated a tremendous spurt forward with the visionary, “Own the Night”. • Pithiness encourages opportunities for inconsistencies in how Rewards (and Outcome Spaces) are interpreted. To the extent that consistency matters, however, a complex system will benefit from continually developing and espousing more detailed and complete characterizations. • However, in complex-system evolution, characterizations cannot be too refined. New Outcome Spaces may need to be added to the characterizations, or their new possibilities will not be 11 explored. See Notes Page
Recommend
More recommend
