ESD.71 Engineering Systems for AnalysisReal Options for Integrating Unmanned Real Options for Integrating Unmanned Aircraft into the National Airspace SystemAircraft into the National Airspace SystemLuke Cropsey30 November 2007Graphic from MIT/LL briefing provided to 303 AESW/XR, May 2007.12/6/2007- ESD.71Slide 2BackgroundBackground Growing need for Department of Defense to fly unmanned aircraft for training and operations within the national airspace system Unmanned aircraft need to be further developed to meet regulatory requirements for integrated flight Significant uncertainty exists in both the cost and performance needed to achieve integrated operations This analysis seeks to address the impact of this uncertainty on the system design12/6/2007- ESD.71Slide 3Options DevelopmentOptions Development Option 1: Pursue a top-down engineering design with significant upfront effortª Intensive requirements definition phaseª Attempt to proceed directly from no access to significant access Option 2: Employ an empirical approach by fielding as rapidly as possibleª Puts out the current state-of-the-engineering as quickly as possibleª Iterates on the design as quickly as possible, feeding results from previous interation into the next Option 3: Flexible framework that seeks to field system quickly, but with foresight towards later requirementsª Seeks to integrate a design for immediate employment with an approach for future extensibility to minimize “switching” costs12/6/2007- ESD.71Slide 4Uncertainty DevelopmentUncertainty Development Probability of FAA approval for unmanned aircraft operationsª Treated as an exogenous system variableª Is path dependent: the way you get there mattersª Will use decision tree analysis to evaluate impact Uncertainty in potential revenue baseª Manned aircraft operations serve as the mission unmanned aircraft will “replace” with corresponding revenue swapª Cost of manned operations is not constant, and may vary dramatically over the develop phase of the unmanned aircraft capability12/6/2007- ESD.71Slide 5Decision Tree AnalysisDecision Tree Analysis Each option evaluated over two decision periods Decision period equal to 5 years cycle time UAS NRE costs only12/6/2007- ESD.71Slide 6Uncertainty DevelopmentUncertainty DevelopmentChange in Manned Operating Costs Probability Distribution with Time0.000.100.200.300.400.500.600.700.800.901.000 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000Dollars (millions)Probability5101520253035404550EV5EV25EV45Change in Manned Operating Costs Probability Distribution with Time0.000.100.200.300.400.500.600.700.800.901.000 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000Dollars (millions)Probability5101520253035404550EV5EV25EV45Probability Distribution of Manned Operating Costs over Time12/6/2007- ESD.71Slide 7Uncertainty DevelopmentUncertainty DevelopmentExpected Value over Time Expected Change in Manned Operations Cost over Time(Includes Maximum and Minimum Values from the same Probability Distribution)02000400060008000100001200014000160000 102030405060Time PeriodsDollars (miilions)MAXMINEVExpected Change in Manned Operations Cost over Time(Includes Maximum and Minimum Values from the same Probability Distribution)02000400060008000100001200014000160000 102030405060Time PeriodsDollars (miilions)MAXMINEV12/6/2007- ESD.71Slide 8NPV with UncertaintyNPV with Uncertainty12/6/2007- ESD.71Slide 9NPV with Uncertainty and FlexibilityNPV with Uncertainty and Flexibility12/6/2007- ESD.71Slide 10Conclusions Conclusions ––UAS Specific FindingsUAS Specific Findings The potential value of addressing uncertainty in key design factors is substantial The AF should approach the problem in a parametric sense--which set of options provides the best investment over what range of values The AF should give specific thought to the utility gained from less than total access and quantify the operational utility of additional flexibility12/6/2007- ESD.71Slide 11Conclusions Conclusions ––UAS Specific FindingsUAS Specific Findings A sensitivity analysis should be performed to focus on a subset of high impact variables to investigateª The high complexity of the effort suggests the analysis would benefit from applying Wang’s option identification process to narrow the trade spaceª Analysis on R&D costs and utility of partial access should be done to determine how well they need to be characterized for a reasonable answer Discussions with the FAA on how this approach could impact flight authorization should be accomplished as soon as possible12/6/2007- ESD.71Slide 12Conclusions Conclusions ––General ObservationsGeneral Observations Different types of uncertainties require different approaches and tools Even a small amount of information with reasonable assumptions about the uncertainty can provide significant insights into design values A synthesis of the decision tree and the binomial lattice approaches into a single analysis tool would provide significant benefit for this kind of option
View Full Document
Unlocking...