Decision Support Tools for Safety Research Investments

Transcription

1 Decision Support Tools for Safety Research Investments Mary Reveley, Program Assessment Team NASA-Glenn Research Center National Aeronautics and Space Administration Fourth National Workshop on Risk Analysis and Safety Performance Measurements in Aviation August 29, 2002

2 Outline AvSP Program AvSP Program Assessment Metrics Decision Support Tools Objectives Decision Support Tool Research Concluding Remarks

3 AvSP Organization Program Office George Finelli, Acting Director Brian Smith, Dep Prog Mgr (ARC) Ron Colantonio, Acting, Dep Prog Mgr (GRC) Glenn Bond, Senior Prog Analyst Technical Integration Frank Jones (LaRC) Program Integration Michael Basehore (FAA) Carrie Walker (Hq)

4 AvSP Projects Aviation System Monitoring & Modeling Dr. Irv Statler (ARC) System-Wide Accident Prevention Dr. Bettina Beard (ARC) Single Aircraft Accident Prevention John White (LaRC) System Monitoring Data Sharing Data Analysis Human Error Modeling Maintenance Human Factors Training Program Human Factors Health Management & Flight Critical Systems Design Control Upset Management Engine Containment

5 AvSP Projects Weather Accident Prevention Gus Martzaklis l (GRC) Accident Mitigation Bob McKnight (GRC) Synthetic Vision Daniel Baize (LaRC) Aviation Weather Info Communications Turbulence Systems Approach to Crashworthiness Fire Prevention Commercial & Business Aircraft GA & Rotorcraft Enabling Technologies

6 AvSP Program Assessment Metrics Safety benefits accident rate system risk Technical Development Risk Implementation Risk Cost/Return on Investment

7 Decision Support Tool Objectives Enable measurement of technology benefit as an effect on risk Enable the translation of this effect measurement into monetary units Provide a process that produces results that are: reproducible, objective, and quantitative Provide a process that describes the uncertainties associated with the results

8 Decision Support Tool Research ACCIDENT MODEL SCENARIO HAZARD CONTROLS envirnmt sys design ops human CONSEQUENCE less severe severe ACCIDENT $$ Hazard elimination Hazard reduction Hazard control Hazard effect mitigation RISK is a function of (1) likelihood of hazard occurring, (2) likelihood hazard leads to an accident (duration and exposure), (3) severity of consequences of the accident

9 PRA Model for Synthetic Vision & Turbulence Detection Technologies Technology Review: System Definition, Planned Capability, Failure Mode Identification Infrastructure models System requirements System design & maintenance policies Human Factors Models Operations models Flight operations (e.g., POAGG) Weather/hazard/ traffic models Reliability model Interaction - response models Integrated safety analysis Infrastructure ilities metrics Operations safety metrics P(Resultant Hazard) = P(Component Failure) * P(Accident Failure) Led by Mary Reveley (GRC) with support from the Logistics Management Institute

10 PRA Model for Maintenance Error Process Description Maintenance frequency and error rate Accident/delay cause and consequences estimate Maintenance human factors effectiveness and benefits Led by Mary Reveley with support from DOT/Volpe

11 PRA Model for Maintenance Error Flight Frequency (F/hr) FAA Airline Statistics Delay/Depart On time (ETMS) Maintenance Frequency (MF) Calculated Based Upon Airline Practice Maintenance Error and Frequency Rate Maintenance Error (ME) via FTA Estimated From: SPAS Datasets Accident/Incident Reports Maintenance Delay Frequency& Duration data Error Consequences (Accident/ Incident Delay) Calculated: (F/hr) x (MF) x (ME) x (Consequence) = Maintenance Error/Flight Hour Cost (Unweighted by Causal Attribution) Accident Cause ASAFE Causal Analysis Spreadsheets (FAA/NTSB Data) Weight-to- Maintenance Cause Assigned Based Upon Accident Report Evaluation Benefits ($ s per Error/ Accident Avoided) Calculated: (Weight) x (Consequence) Accident/Delay Cause and Consequences Estimate Break-Even Benefits Point via BM HFACS Causes Description of Accident Causal Factors (HFACS) Assigned to Individual Accident Reports (Used in Error Rate Calculation) NASA Product Effect (HFACS Causes) Expected Effect of SWAP Program Products by HFACS Category (Based Upon SWAP Documents, Program Descriptions, etc.) Effectiveness Rate Range Allowed to Vary from % SWAP Program Effectiveness and Benefits

12 Probabilistic Decision Support to Evaluate Technology Insertion Objective: To Provide a prototype capability that demonstrates the effectiveness of risk mitigation strategies, such as technology insertions/interventions in the NAS. Approach: Describe scenarios, identify nodes, construct influence diagrams, build BBN, insert technology/intervention, assess relative risks. Led by Sharon Monica Jones (LAC) with support from Rutgers University

13 Concluding Remarks Tools are being developed to determine the relative safety benefits of inserting new technologies into the NAS. Cost/ROI assessments also needed to make new technology insertions attractive to end users