Clean Sheet Data Model for Aviation Safety: Briefing to OMG Transportation Domain Task Force: Clean Sheet Data Model for Aviation Safety: Briefing to OMG Transportation Domain Task Force Rick Jordan, Volpe National Transportation Systems Center
U.S. Dept. of Transportation/Research & Special Programs Administration
Jan Popiel, CSC Corporation
7/10/2001 Danvers, MA Telephone: 617-494-2818
internet email: jordan@volpe.dot.gov
Agenda: Agenda Overview of the Task
Data Difficulties
Data Model Products
Safety Data Events
Hazard & Risk model
Air Operator Organization, Personnel & Training model
Aircraft Identification, Categorization & Maintenance model
Flight model
Conclusion
Part I - Overview Aviation Safety “Clean Sheet” Data Model: Part I - Overview Aviation Safety “Clean Sheet” Data Model What is it?
A formal set of data models and business rules being developed to promote a common understanding of the domain of aviation safety data
Contains data entities from most parts of the aviation business domain
“Clean”
Unconstrained by today’s data structures and legacy practices
Includes some structures not yet automated
Part of FAA response to White House Commission on Aviation Safety & Security - 1997
Clean Sheet Objectives: Clean Sheet Objectives Provides structures for system safety analysis
Articulates a rationale for structuring aviation safety data for proactive risk assessment approach
Provide a roadmap for expansion and renovation of aviation safety information and systems
Supports planning for tools and systems through data requirements and design implications
Promotes data sharing through standard data & references (taxonomies)
Defines the relationship of source information to performance metrics
Clean Sheet Customers: Clean Sheet Customers Customers Who Have Benefited
FAA Office of System Safety - Concepts, definitions, terms, business rules
FAA Systems Builders
Research Community engaged by FAA
International & external aviation data system managers – CAST/ICAO, NTSB
Customers Who Might Benefit
Other parts of the FAA
Handbook & Regulation Writers - definitions
European & Canadian Civil Aviation Authorities (CAAs)
Airlines
Aviation Clean Sheet Data Model Planned Subsets : Aviation Clean Sheet Data Model Planned Subsets Hazard and Risk
Air Operator Organization, Personnel & Training
Inspections & Performance
Aircraft Identification, Configuration & Maintenance
Flight
Landing Facility Subset
Airspace System Subset
Program Schedule: Program Schedule Month/Year Milestone of the Clean Sheet Data Model for Aviation Safety Jun-99 Started work Aug-99 Initial data model for proposed Aircraft Identification Standard Nov-99 First version of Hazard/Risk Data Model May-00 Revised version of Risk/Hazard model - May 2000 Dec-00 Organization, Personnel & Training Data Model Oct-00 First Version of Aircraft Categorization & Identification Data Model Jan-01 Aircraft Identification, Configuration & Maintenance Data Model May-01 Flight Data Model Aug-01 Landing Facility Data Model Sep-01 Airspace Data Model
Slide8: Information System Life Cycle Phase/Stage Data
Model Initiation
Phase Concept
Phase Requirements Stage Design
Stage Conceptual
Data
Model Logical
Data
Model Physical
Data
Base
Design Initial Data Entities Scope
High-level Data
Entities Refine Scope
and Create
Conceptual
Data Model Develop
Physical Data
Base Design Logical Data
Model with Data
Elements Revised
Logical
Data Model Add Data Entities
and Data Elements
From Process/
Data Analyses Revise Logical
Data Model
Based on
Normalization Logical
Data Model Planning
Sources: Sources FAA documentation
FAA orders, procedures & policies, regulations, Handbooks, Manuals, etc.
National Airspace System (NAS) information
System documentation and data dictionaries
Research and Published Information
System Safety texts - Includes Wil Hammer & MIL-STD-882D
FAA Technical Center Research publications
Interviews and Working sessions with FAA staff engaged in aviation safety analysis
IT Materials and Best Practices - generic data models
Aviation Industry Documentation
Flight Safety Foundation
Eurocontrol information
Aviation safety web sites - RTCA, NASA, GAIN, etc.
CAST/ICAO Common Taxonomy Team
Agenda: Agenda Overview of the Task
Data Difficulties
Data Model Products
Safety Data Events
Hazard & Risk model
Air Operator Organization, Personnel & Training model
Aircraft Identification, Categorization & Maintenance model
Flight model
Conclusion
Attributes of Quality Transportation information: Attributes of Quality Transportation information Timeliness: up-to-date
Accuracy
Correct data definitions – unambiguous, clear, consistent
Definitions are cross-functional for operational, tactical & strategic functions
Completeness – missing values
Relevant
Obtainable
Format and presentation of data is appropriate, understandable & re-useable Items in bold benefit from data modeling
Data Difficulties: Data Difficulties Difficulty having corporate definitions and unique identifiers of key entities like:
Aircraft, Aircraft groupings, Facilities
Airline Identifiers: FAA and BTS lists
Redundant updating of lookup tables (reference tables) by 2 or more systems & across agencies
Strategic & Tactical functions need:
Compatible & quality data from many operational systems
Ability to aggregate or summarize data from operational systems
Impacted Tactical functions include:
Aviation Safety Analysis: Incident Rate as a Percent of an Airline’s Operations
Relating the work/outputs of CAAs to desired safety outcomes
Estimating costs across the life cycle of an asset – How much have we spent on System Z since we did the research, acquired it and have had it installed for the past 3 years?
Airbus A300 - Make/Model/Series: Airbus A300 - Make/Model/Series FAA Aircraft Registry National Transportation
Safety Board Airclaims (UK, private) Service Difficulty
Reporting System Near Mid-Air Collision System Same System = Different Formats Different Systems = Different Formats As of 6/99
Agenda: Agenda Overview of the Task
Data Difficulties
Data Model Products
Safety Data Events
Hazard & Risk model
Air Operator Organization, Personnel & Training model
Aircraft Identification, Categorization & Maintenance model
Flight model
Conclusion
Safety Event (accident, incident, etc.) Data Conceptual data model of circumstances and precursors: Safety Event (accident, incident, etc.) Data Conceptual data model of circumstances and precursors Aircraft / Other Asset
(event’s primary subject)
e.g. aircraft instance, ATC asset Safety Event
e.g. Accident, Incident
- What happened? (big picture category)
- When it happened? (service,phase,flight)
Where it happened? (coordinates)
Who it happened to? (operator)
What aircraft (model) it happened to?
(represents actual unplanned safety events or hypothetical event scenarios) Event Category
e.g. runway incursion, CFIT, … Organization / Person
and air carrier system elements (e.g. maintenance personnel) Phase of Flight Event Factor (events tree,
causal factors, precursor hazards)
- what aircraft/asset component failed
what Failure Mode, Factor Category it is
what operation / maintenance task it is
who was involved (role rather than name)
how it happened (preceding events)
why (causal/contributing factor, hazard) Airport / ATC Facility Jan Popiel, May 29, 2001 of what subject when what where related to comprises Aircraft / Asset Type
aircraft model (e.g. Boeing 747-200C)
engine type/model (e.g. GE XY)
other aircraft equipment / component
ATC equipment type/model Hazard / Risk Factor
Category
human factor category
technical failure by ATA code
external environment (weather) is a is a how & why Role / Responsibility
pilot, controller, inspector, mechanic, … who who who Reference Data
(Dimensions & Taxonomies) What events happen? When events happen? Where events happen? operator for How & Why did events happen? what object related to what object involved in Failure Mode
& Effects
- severity
likelihood / rate
failure effects of is a Airspace Class categorizes codifies location of involved in Who was involved? What hazards exist? Airspace Structure
Sector, Route, Fix, … embodied in How a thing /
component breaks? where Safety Rule / Standard
e.g. Visual Flight Rules concerns What controls/standards exist? comprises Aircraft / Other Asset Reference Data Operation/Service Type
e.g. commercial, GA, … has involved in involved in has What tasks
prevent failures? Flight of has Preventive
Maintenance Requirement
/ Task (?)
task frequency
Notes:: Notes: Reference category data Color legend: Safety event data Aircraft / asset data Organization /Person data Aspects not covered:
Which database/system the event or factor data came from.
Safety Event Data: Safety Event Data Definitions
Safety Event– is an unplanned and undesired aviation accident or incident. It may exist for an aviation asset or aviation asset type/model
Event Factor (includes hazard) – is a state, condition, or action that occurred during a Safety Event. It includes all events in a string making up an unplanned safety event – including corrective, negative, and normal aspects.
Shows the “dimensions” of a safety event – the who, what, when, where, why & how
Some of these dimensions are subject to standardization efforts for a common taxonomy or standard reference list of values
Indicates the relationship & scope of each taxonomy effort
Generalize-able to any safety event – applicable to aviation and other domains
Hazard / Risk Data Model: Core Data Entities for Hazard Analysis & Risk Assessment: Hazard / Risk Data Model: Core Data Entities for Hazard Analysis & Risk Assessment Safety
Event Event
Factor
Hazard & Risk Data Model Value Added: Seven Fundamental Entities : Hazard & Risk Data Model Value Added: Seven Fundamental Entities Safety/Unplanned Event: Accident, Incident, Hazard Scenario
Includes scenarios for hazard identification
Event Factor: Hazard, Precursor, Corrective action, Outcome, normal actions
Event Factor Association: Causal Chain
Risk Evaluation / Rating
Likelihood & Severity
Risk Mitigation Strategy: Safeguard, Safety Defense
Event Taxonomy
Hazard / Risk Factor Taxonomy
Application: Incident: Application: Incident NTSB Identification: DCA991A019
Scheduled 14 CFR 121 operation of DELTA AIRLINES
Incident occurred DEC-06-98 at NEAR HAMPTON, NY
Aircraft: Boeing 767, registration: UNK
Injuries: 0
On December 6, 1998, about 2200 eastern standard time, Delta Airlines flight 66, a Boeing 767, and Caledonian Airways flight 5199 (CKT5199), a Lockheed L1011, were involved in a near midair collision approximately 10 miles south of Hampton, New York at 33,000 feet (FL330) mean sea level. Both flights were being handled by the Boston Air Route Traffic Control Center (Boston Center) According to the FAA, at 2145 the Boston Center computer display channel (CDC) failed and the direct access radar channel (DARC) equipment loss its ability to interface with the National Airspace System (NAS). Because of this, most of the aircraft identification, track, and flight plan data normally displayed on controllers’ radar displays, were unavailable. (Transponder beacon codes and Mode C altitude information continued to be displayed.) The radar controller accepted a handoff from New York Center on DAL66 level at FL330. The same radar controller then accepted a radar handoff on CKT5199 Level at FL330. The flights were on converging courses. The flight crew of DAL66 received a traffic alert and collision avoidance system (TCAS) resolution alert (RA) to descend, and the Flight crew of CKT5199 received a TCAS RA to climb. At 0300 UTC, the two airplanes passed each other with 1.5 miles horizontal and 900 feet vertical separation. The pilot of CKT5199 reported sighting DAL66 and said he would file a near midair condition report.
Event Factors for Near Mid-Air Collision Delta 66: Event Factors for Near Mid-Air Collision Delta 66 11. Delta 66
Approach 4.
TCAS
Alert 9. DARC
Capability
Loss 8. CDC
Failure 10. Caledonian
5199
Approach 1. Near
Mid-Air
Collision 5. Loss Of
Separation 6. Incorrect
Navigation
Directions 7. Loss of
Radar
Display 2.Evasive
Action
(Climb) 3.Evasive
Action
(Descend)
Translation into Data Tables: Translation into Data Tables
Air Operator Organization, Personnel & Training Data Model – Scope & Sample Entities: Air Operator Organization, Personnel & Training Data Model – Scope & Sample Entities Organization
Org Relationship
Person
Aviation Involvement Certificate
Certificate type
Designator
Aviation Involvement Aircraft Id Grouping
Grouping Relationship
Component
Aircraft Instance Training
Course
Task/Skill Type
Course Topic
Airmen Certificate
Position Type Inspection
Inspection Plan
Inspection Elements
Insp. Results
Action Taken Performance
Safety goal
Performance Measure
Performance Target
Performance Value
Safety Taxonomies Safety Work Activity
Certification
Inspection
Safety Activity Report
Air Operator Organization, Personnel & Training Data Model Value Added: Refined Basic Concepts/Entities: Air Operator Organization, Personnel & Training Data Model Value Added: Refined Basic Concepts/Entities Organization: currently tracked via FAA’s Designator but needs to be more general than Air Operator and Agency
Organization Involvement: allows multiple roles in aviation
Organization Relationship: identifies relationships between Orgs.
Hazard/Risk Taxonomies: need mapping for aggregation/analysis
Many taxonomies: ACAT, ICAO, ATOS Air carrier system
Related taxonomies to other safety factor classification schemes
Inspection Plan: generalized for any inspection type
Safety Criticality Ranking: modeled to be used both for driving inspection frequencies and implementation
Inspection Conclusion: drives corrective activities
Aircraft Identification Configuration & Maintenance Data model - Approach: Aircraft Identification Configuration & Maintenance Data model - Approach Partly done for CAST/ICAO Common Taxonomy Team on Aircraft
Seeking a standard reference list of aircraft make, model, series
Configuration control (Aircraft components such as engines)
For individual aircraft (aircraft instances) and
For asset types (such as an aircraft make)
Minimum equipment list (MEL)
Aircraft maintenance
Maintenance task and their resource requirements
Reliability and failure modes
Aircraft Asset Data Model Fundamental Entities: Aircraft Asset Data Model Fundamental Entities Aviation Asset
Aircraft
Airspace System Asset
Aircraft Maintenance Asset
Aviation Training Asset
Other Aviation System Asset
Asset Type / Model
Aircraft Identification Grouping
Engine Type
Other Asset Type Maintenance
Program
Task Type/Procedure Definition
Task Resource Requirement
Operation/Support Task
Documents used (such as maintenance manuals)
Failure Mode/Anomaly
Equipment Type List
Slide27: Aircraft Categorization and Identification Aircraft Identification Grouping
# Grouping Id
Aircraft Make-Model-Series -Manufacturer Grouping
* Make
o Model
o Series
o Manufacturer
o Aircraft Popular Name
o Amateur Built Flag
Aircraft Instance
# Aircraft Instance Id
* MMS Grouping Id (FK)
* Make (derived)
* Model (derived)
* Serial Number
* Latest Registration Number
* Effective From Date
o Effective Until Date
(and other attributes) Type Certificate
Grouping
Aircraft Grouping
Association
# Grouping Association Id
* Grouping Association Type
* Source Grouping Id (FK)
* Target Grouping Id (FK)
o Remarks Organization
# Organization Id
* Full Name
* Short Name
* Country Code
* Effective From Date
o Effective Until Date Aircraft Instance Other Categorization
# Grouping Id (FK)
# Aircraft Instance Id (FK)
o Remarks Aircraft Grouping Structure Component
o CNS Equipment Suffix Code Aircraft Instance
Structure Component
# Component Instance ID Aircraft Instance
Legacy Identification
# Aircraft Instance1 Id (FK)
# Aircraft Instance2 Id (FK)
o Remarks Component Type/Model
source target Conceptual data model
April 5, 2001 Organization Aircraft Involvement
# Involvement Id
* Role Type Code (function)
* Effective From Date
o Effective Until Date
* Involvement Country Code
o Remarks comprises comprises comprises comprises model for instance of was is embodied in embodied in is Aircraft Primary Characteristic
* Aircraft Category
o Aircraft Sub-Category
* Number&Type of Engines
o Unregulated Vehicle Type
* Weight Category
* Noise Level Stage
* Landing Cert Category
* Max Cert Takeoff Weight
* Median Climb Rate
* Median Descent Rate
* En Route Speed Range
* Take off Speed Range
* Approach Speed Range
? Passenger/Cargo Flag
has e.g. equivalent, parent/child, characterized by involved in involves for of involves Aircraft Type Designator
Grouping
* Type Designator
* Climb Rate
* Descent Rate Landing Gear Model Engine Model CNS Equipment Model Other Equipment Type new/old, preliminary/final Other Aircraft Type
Grouping
Organization Association
# Org Association Id
* Association Type
* Source Org Id(FK)
* Target Org Id(FK)
o Remarks out of scope of the initial development phase manufactured by for for is is of
Value Added: Selected Highlights: Value Added: Selected Highlights Relationships to Failure Mode or Condition Type
Business Rule: When an asset defect or condition is observed, then whenever possible the standardized failure mode or condition type should be identified (as its cause).
Business Rule: A maintenance requirement may be defined for a particular failure mode or condition type. If a condition is met, then the maintenance steps need to be performed.
Business Rule: A maintenance requirement should reference the aviation product category such as ATA Specification 100 code to identify a category of components that are maintained.
All of the above supports analysis of past history for safety issues
Facilitates aggregation of aviation safety information into different aircraft categories
Flight Data Model: Flight Data Model Incorporation of Flight event entities into the Clean Sheet Data Model
Unplanned Events & Event Factors
Risk Reduction/Corrective Measures
Organizations
Asset Type/Model
ETMS information (tactical air traffic control system)
Flight data recorder information
Value Added: Selected Highlights: Value Added: Selected Highlights Relationship between Flight Event and Event Factor
There is a many to many relationship between Flight Events and Event Factors.
Many flight events can be the source information for one event factor
An event factor can involve several flight events because a safety analysis may be performed multiple times using different risk / hazard analysis methods
Collaborative Information Exchanges can take place concerning an Unplanned Event (usually an observed air traffic control problem).
A collaborative information exchange can result in a Safety Work Activity (such as further Risk Analysis Activity).
A CDM Problem Identification Notice can be relevant to an Aviation Performance Metric Value.
Landing Facility Data Model: Landing Facility Data Model Underway through August
Sources:
Airport Certification Handbook
Airport Master Records including changes to airports, runways, & facilities
Airport Improvement Handbook
ARP organizational information & goals
Various ARP forms
ARP Policies, Orders, etc.
Runway Safety Area Inventory
System Requirements document(s) for ARP systems
Any other suggested sources?
Direction : Direction Planned work on Airspace Data Model: Aug/Sept.
Other Coordination and Communication Steps Information on our Data Modeling work can be found at:
http://www.volpe.dot.gov/data/info.html
Part III - Conclusion: Part III - Conclusion Provides structures for system safety analysis
Articulates a rationale for structuring aviation safety data for proactive risk assessment approach
Provide a roadmap for expansion and renovation of aviation safety information and systems
Supports planning for tools and systems through data requirements and design implications
Promotes data sharing through standard data & references (taxonomies)
Defines the relationship of source information to performance metrics
Contact
Rick Jordan, Volpe Center
Jan Popiel, CSC
Appendix: Appendix Introduction to Data Modeling Concepts and Graphical Notation
Basic Concepts of Data Modeling: Basic Concepts of Data Modeling Entity
A person, place, thing, concept, or event about which an organization wishes to collect, keep & process information
Relationship
An association between two entities established to represent their structural, causal, and integrity dependencies
Attribute
An item of information describing and entity (it’s property)
Business rules mostly represented by relationships
Simple Example: Simple Example Two entities: Flight and Aircraft
Relationships:
A Flight is scheduled for one aircraft
An Aircraft flies many flights Flight
Aircraft
How to read an Entity-Relationship (E-R) diagram: How to read an Entity-Relationship (E-R) diagram Program Number
Program Name
Program Description Project Number
Project Name
Project Description
Approval Date Entity Name
Attributes
Relationships Each entity has the potential to be implemented physically as a database table.
Each attribute can become a physical data element. Line Item No.
Sub Line Item No
Amount Line Item No.
Amount one
many Program Project Underlined Attributes
are potential primary keys Budget Line Budget Sub-line