Presentation on theme: "The Value of Technical Performance Measures"— Presentation transcript:
1 The Value of Technical Performance Measures Ann LynnThe Boeing Co.(314)All Data in this pitchis Notional – for example only
2 Agenda TPM Definition Typical Types of TPMs Constructing TPM Profiles Performance ThresholdsPlan Over TimeUncertaintyManagement ProcessAllocation of MarginRoll-up (Parent/Child) TPMsValue to a Program
3 Technical Performance Measures Represent critical technical thresholds and goals for success of the programShould be variable and responsive to engineering changesUsually exist in a hierarchy corresponding to the spec levels (system, subsystem, component…)Have target values corresponding to specification valuesthat is, your TPM should be a parameter in a spec at some levelAre tracked and statused against their defined plan and represent the technical health of the program
4 Typical Types of TPMs Parameters that flow from and/or support KPPs Key Constraints and Performance ReqmtsWeight, power, heat loadReliability/ Maintenance parametersSystem / subsystem performanceEnduranceRangeLatencyAccuracyEfficiencyEtcCost (e.g. AUPC)Parameters associated with Program Technical Risks or areas of technical challengeTPMs should represent a balanced set of key parameters;Don’t monitor something just because “you can”
5 Example TPM Status Chart Example Monthly / Quarterly StatusThis view shows: •Title/Definition •Status against spec•Status Color (RYGB) •Maturity of the status
6 Definition of Performance Thresholds (Generic) Blue – the system exceeds requirements by more than x (and may be subject to re-allocation)Green – the system meets all performance requirementsYellow – the system can meet some objectives but does not meet all performance requirementsRed – the system cannot perform satisfactorilyTPM marginSpec valueTPM marginThis is the standard method for defining TPMs. Note that the margin below the spec takes the paramter from “meets” (green) down to the bottom of the yellow zone. That way, when a parameter is not meeting its spec, you ask “how much margin is there?” meaning how much “yellow zone” do you have before that parameter hits red. Where yellow zone is that area where you can negotiate yoru allocated requierments to try to still achieve objectives & goals of the program. Typically setting the spec value to the “hairy edge” of needed performance is considered too risky and doesn’t’ allow the opportunity to use the TPM process Mechanisms for the yellow and red zones.Note, reqmts can be negotiated/updated withcorresponding updates to Specs and TPMs.Setup as shown, the parameter has to meet the spec to be green.
7 TPM Profile Over TimeThis view of the TPM allows you to see history and trends in performance
8 Profiles over time (cont.) TPMSpecStatusUnit Efficiency24%22.2%(Meas.)In this example, the team receives samples from the supplier and expects incremental improvement over time to eventually achieve the spec value. If each sample comes in at its target value, the TPM status should be Green. Therefore the green range changes over time in accordance with the plan. TPMs are set up as “performance to plan”. Note that this type of graph is not intended to reflect changes to spec requirements. TPMs are spec-based, so if the spec changes then the spec line would be adjusted and the threshold values redone so that spec is still the bottom of the green.TPM using current status, not predicted status at end of programFor this TPM, performance is planned to improve over time.If actual progress >= planned, you are green, even if you haven’t achieved the end-state target (spec value)
9 Weight ManagementProjected weight (projected value at end of program)includes a historical growth factorCurrentProjected Wt.(w/ Growth Risk)lbSpecNTE Weight3850 lbTPM if using current statusTPMSpecStatusSystem Weight38503588 (current)IPT WeightAllocations3588 lbCurrentBaseline Wt.lbTPM if using projected statusBaseline weight reflects current configuration sized to loads at NTE weight requirement.Projected weight includes planned design changes and X% weight maturity growth risk for post PDR design maturity.TPMSpecStatusSystem Weight38503971(projected)Weight TPM usually plotted per the Weight Mgmt process;Show both values, status to Projected;99
10 TPM Profile with uncertainty bar include an “uncertainty” bar for each reported valueSpec value lineIf your project has significant Risk, or it is early in the project,TPM status can be deceiving unless uncertainty is understood
11 TPM Management Process (Generic) In the blue: consider adjusting that parameter’s spec value and taking relief elsewhereIn the green: press on!TPM marginSpec valueIn the yellow, or alarming trend: take a program-defined action (put on Risk Watch list, initiate reqmts/design trade, establish new risk mitigation, initiate change to design, reqmts, or plan)TPM marginThese process mechanisms allow the program to define set actions based on the severity of concern regarding the status.In the red: take a program-defined action (establish new risk mitigation, initiate change to design, reqmts, or plan)The approach to monitor and control is predicated oncorrect allocations of margin
13 Definition of TPM Margins Factors to consider:Margin may be allocated based on the amount of risk (e.g. Red Risk or low TRL subsystem gets a greater share of margin)Margin may be allocated based on relative impact to System performance (accounts for sensitivities)Margin may be limited by the “hard points” - those parameters that have a hard stop which prevents acceptable operation (e.g. structural limits for weight)TPM marginSpec valueTPM marginMargin: How bad can it get before status goes red?How good must it be before status goes blue?
14 *System Parameter* Tree Power ConsumptionEnergy CollectionSPC efficiencyPropulsion ConsumptionSubsys Pwr ConsumpEnergy StorageVMS / MSProp efficiencyAV WeightCruise consumpSubsys ZAero efficiencyVMSESS RT efficiencymaneuversSubsys ZESS componentsDragA componentsESS degradationStabilityESSB componentsESS paramPower DistributionSubsys ASubsys BEPD efficiencyPropulsionPPX efficiencySubsys XConvertor efficiencyHigh SensitivityLower sensitivitiesThis diagram maps the subsystem parametersthat contribute to the system-level parameter;Sensitivities are noted.Wiring efficiency
15 Associated Risks *Sys Parameter* This diagram maps Risk to Power ConsumptionEnergy Collection46-xxxSPC efficiencyPropulsion ConsumptionAvionics & PSS Consump24-xxxEnergy StorageVMSProp efficiency1-xxxAV WeightCruise consumpSubsys ZAero efficiencyVMS57-xxxxESS RT efficiency41 - xxxmaneuversSubsys ZESS componentsDrag54-xxx7-xxx18-xxxA componentsESS degradation51-xxxStabilityESS6-xxxB components50-xxxESS paramPower DistributionSubsys A2-xxxSubsys BShows Risk #, Title, and current level (RYG)EPD efficiencyPropulsion56-xxxxPPX efficiencySubsys X56-xxxxConvertor efficiencyHigh SensitivityLower sensitivitiesSubsys x paramWiring efficiencyThis diagram maps Risk tothe subsystem parameters.Helps show where margin is needed.
16 Interdependent TPMs Competing for their share of the overall margin Their performance needs to be monitored as a groupChange (re-allocation of margin, spec adjustment) has to be managed for the group
17 KPPs and TPMs KPPs TPMs System level Sub- & Cmpnt Time of Yearat which Energy balanceis closedTPMsSystemlevelSub-&Cmpnt*System Parameter*(AV Efficiencyat End of Mission,AV Performance)AV WeightAV Power ConsumptionPropulsion effESS round trip effPPX effSPC effAV Aero effPropulsion Power consumpVMS Power consumpAirframe Power consumpESS Power consumpSubsys A Power consumpSubsys Z Power consumpVMS weightPropulsion weight-Airframe weightESS weightPPX weightSubsys A weight-Subsys Z weightKPPs represent customer operational objectives;TPMs shown represent measurable design parameters that contribute to the KPP
18 Relationship of System and subsystem TPMs Example: Command Latency <= 160 msIf supplier specs are more stringent, that means margin is held at subsystem level above supplierSpreadsheets can help account for values and margins at each levelMany PMs prefer to monitor Supplier-owned TPMs when availableSystemmarginWeaponLnchrSMSMCPVI25525255030PVI CMPNT1Spec =35 msWeaponSpec =20 msLauncherSpec =5 msPVI CMPNT2Spec =15 ms
19 Example parent/child TPMs (1 of 3) OwnerTPM TitleSpec ValueCurrent StatusCESystem Command Latency160 ms136 msDsplys^PVI Cmd Latency50 msMsn Cmptr^MC Cmd Latency25 msStores^SMS Cmd Latency^Lnchr Cmd Latency5 ms^Weapon Cmd Latency31 msSystem TPM can be green even when child (subsystem) TPM is red;in this case because of margin held at system level
20 Example parent/child TPMs (2 of 3) OwnerTPM TitleSpec ValueCurrent StatusCESystem Command Latency160 ms121 msDsplys^PVI Cmd Latency50 ms35 msMsn Cmptr^MC Cmd Latency25 msStores^SMS Cmd Latency^Lnchr Cmd Latency5 ms^Weapon Cmd Latency31 msSystem TPM status can be green if child (subsystem) TPMs balance out
21 Example parent/child TPMs (3 of 3) Example: Command Latency <= 160 msSystemmarginWeaponLnchrSMSMCPVI255252550303520WeaponSpec =25 msLauncherSpec =5 ms35Over time, System TPM margin may need to be re-allocatedto compensate for subsystem over targetSubsystem TPM in the blue range may have spec adjusted and“give back” margin to the System level or to another subsystem
22 TPMs - Takeaways Create a balanced set of key parameters Ensure the full Profile is understood by program technical leadership, even if only the stoplight format is used for periodic reviewsProfile over time current vs projected statusBasis for margin maturity / uncertaintyAssess the risk and sensitivity of parameters to support appropriate margin allocationKnow where your margin is … and isn’t, so you can react to changes in reqmts or in design statusTPMs can be used as a predictor of success … and failure
23 The Value of TPMs Track the “right” things Make sure your TPMs correlate to key technical parametersTrack technical progress compared to plannedYour monitoring system is only as good as the plan & thresholds you set it up withSpend the time upfront to set up correctly!Your monitoring system is no more accurate than the data you put inmaturity / uncertainty of reported data must be evaluatedConisder plotting uncertainty bars on data points to avoid false sense of security