The U.S. Navy`s Aviation Data
Transcription
The U.S. Navy`s Aviation Data
DECKPLATE DECKPLATE: The U.S. Navy’s Aviation Data Management System S CDR Christopher Hammond DECKPLATE Deputy Program Manager Naval Air Systems Command 6.8.4 Lisa Clark & Troy Bekel DECKPLATE Support Spalding Consulting, Inc. Wednesday, June 6, 2012 NAVAIR Public Release #2012-705 Distribution Statement A - "Approved for public release; distribution is unlimited" US Navy Today Personnel >325,000 >50,000 deployed Aircraft Carriers (CVN’s) USS George HW Bush; Ronald Reagan; Abraham Lincoln; George Washington; Dwight D Eisenhower; Enterprise; Nimitz; Truman; Stennis; Roosevelt; Vinson Carrier Air Wings (CVW’s) (H-60; E-2C; F/A-18C/E/F/G; C-2; EA-6B) 5 East coast 4 West coast 1 Forward Deployed (Atsugi Japan) 3 Decision Knowledge Programming for Logistics Analysis and Technical Evaluation Overview: Features: Stakeholders: Aviation logistics data warehouse for the US Navy with 23 years of history data for trend analysis and records reconstruction Near real-time, parallelized data loads, Active Data Warehouse, single, authoritative, web-based data y and reporting p g source for analysis -Commander of Naval Air Force (CNAF) -Fleet Readiness Center (FRC) -Headquarters Marine Corps (HQMC) -Type Commanders (TYCOMs) -Office of the Chief of Naval Operations (OPNAV) -Program Managers (PMAs) DECKPLATE Data Subjects: Flight/Usage Inventory Aircraft Maintenance Configuration Baseline Management Engine Total Asset Visibility Technical Directives Supply Cost 4 Official up-line repository for inventory, maintenance, flight, usage and readiness reporting data Fulfills United States Code Title 10 Section 117 requirements for Defense Readiness Reporting System Measures and reports on the readiness of military forces and the supporting infrastructure Fulfills the U.S. Naval Aviation policy requirement for Aviation Data Warehouse (ADW) Provides a comprehensive view of aircraft, engines and aeronautical components across the Naval Aviation Enterprise (NAE) Enables decision makers with usable, actionable and reliable information 5 NAE Naval Aviation Enterprise (NAE) Mission: To support Naval aviation readiness requirements with transparent, cross-functional processes, which inform risk-balanced decisions 6 NAE & DECKPLATE Enabling decision makers with reliable “near real-time” data: Solving Complex Problems Endorsed by CNAF and HQMC as the primary tool of choice to meet daily, weekly, monthly and congressional data calls NAE decision makers use DECKPLATE for: Quarterly cost war room briefs Readiness data to trend degraders Metrics are consistently briefed through: Current Readiness Cross Functional Team Maintenance &Supply Chain Management The AIR Board Best decisions ensure every dollar is wisely spent and focused on Warfighter requirements in a timely manner, for the greatest return on investment 7 NAE’s Complex Mission: DECKPLATE provides the NAE with the ability to measure and improve logistical processes and associated costs File: NAVAIR Brief 8 NAVAIR 6.8.4 Handling the complexities of big data in Naval Aviation Logistics Information Technology 9 DECKPLATE & Naval Aviation 10 Our Customers 11 Typical Data Sample AIRCRAFT MATERIAL MAINTENANCE PROCESS/ METHODS Operating Repairables I-Level FY11 FY11 CPI Events USN: USMC: Total: >1200 >850 >2050 Pipeline p USN: USMC: Total: >250 >200 >450 NSN: Top 500: 10,500 67% of costs Consumables NSN: Top 10,000: 105,000 87% of costs IP: RFI: BCM: D-Action: 136,216 41,657 91,780 2,779 FRC FY11 Airframes • Scheduled: • ISR: Eng/Mods: Components • Lvl Sked: • PBL: 431 3140 930 USN: USMC: FRC: 37 79 474 FY11 BCM Interdictions IP: Saved: >4000 >$80M 33,133 14,307 As of 30 Jun 2011 12 DECKPLATE Data Flow 13 Leading Indicator Metrics CNAF requires current and trending data relative to specific metrics in order to proactively affect outcomes rather than react to problems Problem Traditional reporting: Does not identify problems as they emerge and unfold Focuses on a single subject area; does not cross subject area boundaries or provide correlation between them Does not allow for ad-hoc, trended data, based on current, near real-time data Does not allow for instant results 14 Leading Indicator Metrics DECKPLATE Solution Understand questions that need to be asked Provide queries and data utilizing an Agile approach Automate approved data aggregation and provide for on-demand retrieval 15 Leading Indicator Metrics Results CNAF has access to near-real-time comprehensive measurements for Navy and Marine Corps aviation maintenance and supply asset visibility CNAF uses this data to compare supply and demand of repairable components and resolve trended supply, and possibly manufacturer shortage issues Metrics allow CNAF to accurately schedule deployable assets assets, aircraft and aircrew Aircraft downtime can be projected and a more accurate schedule can be maintained in order to achieve more accurate readiness accounting 16 Ready For Tasking (RFT) Right aircraft, in the Right configuration, in the Right place, at the Right time, with the Right readiness 17 Ready For Tasking (RFT) Availability, at the right time, of a properly configured aircraft and specific aircraft mission systems Key contributor to the accomplishment of the Fleet Response Plan readiness objectives Provides timely data discovery for key readiness factors, for example: Cannibalization Corrosion RFT “Ready Basic Aircraft” Aircraft with a common minimum configuration to conduct flight operations with necessary communications, flight and safety systems required by regulations 18 DECKPLATE RFT New integrated ad-hoc subject area across: Active and Closed Maintenance Actions More timely, better readiness Supply and Cost from NAVSUP (Naval Supply Systems Command) and DLA (Defense Logistics Agency) MESM (Mission Essential Subsystem Matrix) and EOC (Equipment Operability Codes) Used to identify aircraft subsystems with Partial Mission Capability Finer granularity on more data 19 Daily Readiness Reporting Current Readiness – Essential to Naval Aviation's capability posture Problem Monthly data does not provide analysis for specific periods: Deployment Readiness is not a binary up/down status FMC (Full Mission Capable) PMC (Partially Mission Capable) NMC (Not Mission Capable) Complicated readiness determination Current aircraft status All maintenance and usage over specified time period Sifting through massive data quantities To what do you attribute impacts to readiness? 20 Binning Granular Aircraft Status 15 minute “bins” generate immense amounts of data Rank mission capability to highlight aircraft availability Last 5 years Recreate maintenance activity y over a period of time Reporting timeframes show acceptable performance Fleet Benefits Highly accurate picture of aircraft readiness Clearly identify availability cause: supply or unscheduled maintenance Trending capability data for given time periods 21 Binning Challenges Processing Billions of Records Source Data (5 years) Binning 876,000,000 XRAY Status Bins 4,700 Aircraft 106,000 XRAY Status Messages 31,000,000 Maintenance Steps Results 2,000,000,000 , , , Maintenance Step Bins 10 000 000 10,000,000 Daily Result Records 2,600,000,000 Merge Bins 3,000,000 Flights 246,000,000 Rollup Bins 22 Data Transformation DECKPLATE does a tremendous amount of validation, processing, and correlating of incoming data. Expand in Processing to 803,000,000 g Records Incoming EVERY DAY Needed to maintain 4,400,000,000 Transformation R Records d EVERY DAY 2,000,000,000 Permanent Records During processing, incoming data expands by 550% 23 Parts Life Tracking Baseline Configuration Work Unit Code (WUC) Structure/Framework Maintenance Schedule Allowable Parts/NIIN (also cost) Life Limit (drives maintenance) Actual Configuration Events and Usage for each WUC Bi Data Big D t Challenge Ch ll : Smart Aircraft Sensor captured usage data E.g. F/A-18 engines can have 1,500+ parameters on each engine >2,000 F/A-18 engines captured at various intervals per flight CREATES MASSIVE DATA “Smart Data” expanding to other platforms E.g. H-53 collects 37MB of data per hour of flight time 24 Change Data Capture (CDC) Big Data Refresh Challenges Source systems cannot provide consistent revision date information DECKPLATE must retrieve whole data set each time for comparison to previous data set About 2 billions rows compared each day Comparisons moved from OS file-based to database Massively Parallel Processing (MPP) provides strong performance Code base is simpler, smaller, and easier to maintain Improved error logging and metrics capture Implemented more business rules and flexibility 25 Change Data Capture (CDC) CDC Levels Primary CDC performed on current vs. previous source data SQL used to identify Adds, Changes and Deletes Duplicates filtered Orphans reprocessed Compensates for data purging in the source system Secondary CDC performed on incoming vs. warehouse data Ignores changes where source records are stale vs. warehouse Verifies source system UPSERTs Converts INSERTs to UPDATEs and vice-versa where needed Maintains warehouse integrity when source data has data issues 26 Active Data Warehouse Third Normal Form (3NF), with some: De-normalized query structures Aggregated/derived data Data expansion/explosion, then ranked and merged Star schema/data mart (OLAP) 24x7 online availability of all data to all users 24x7 24 7 lloading di source data d t Transactional (OLTP) Change Data Capture (CDC) Multiple synchronous load streams (~4 at any point in time) Massive data transformations Combination of parallelized as well as serial/procedural loads Predominantly UPSERT with some full data set replace (large reference data) Massively Parallel Processing (MPP) 27 28 DECKPLATE Tomorrow DECKPLATE is incrementally replacing the Naval Aviation Logistics Data Analysis (NALDA) Aviation Data Warehouse and other legacy systems By 2017, a total of 15 applications will have been migrated, with their functionality absorbed by DECKPLATE DECKPLATE is striving to be predictive with the capability of providing future trending and utilizing new and emerging technologies trending, Utilizing smart IT, DECKPLATE eventually will inform the Sailor/Marine of a problem DECKPLATE will reduce manpower requirements needed to analyze massive data from disparate systems to identify emerging issues and discrepancies 29 Summary DECKPLATE is a solid investment for NAVAIR ensuring the big data challenges facing NAE logistics IT are met Total of 15 legacy applications’ functionalities will be absorbed into DECKPLATE Modern avionics provide much more robust key performance data to meet the increasingly complex conditions under which the Navy & Marine Corps must perform Greater data volumes require massive processing capabilities Affecting outcomes: Identifying potential problems before they happen DECKPLATE WILL CONTINUE TO PROVIDE QUALITY READINESS DATA TO THE WARFIGHTER 30 Contact Information CDR Christopher Hammond Deputy Program Manager NAVAIR 6.8.4 301-757-2311 [email protected] Lisa Clark Senior Vice President Spalding Consulting, Inc. 301-737-0150 [email protected] Troy Bekel Vice President Spalding Consulting, Inc. 301-737-0150 [email protected] 31