FAA AAS Federal Aviation Administration Advanced Automation System

  • Slides: 10
Download presentation
FAA AAS Federal Aviation Administration Advanced Automation System • Designed to replace aging equipment

FAA AAS Federal Aviation Administration Advanced Automation System • Designed to replace aging equipment from ‘ 60 s and ‘ 70 s • Also consolidate ~200 tracons into ~20 enroute centers • 1981 -1994, cost billions of $ • Some parts of new equipment (controller workstations) used, but system as a whole failed • Consolidation idea abandoned

FAA AAS Requirements • • • Distributed system 99. 99999% availability (3 sec/year downtime)

FAA AAS Requirements • • • Distributed system 99. 99999% availability (3 sec/year downtime) No paper (electronic flight slips) New hardware New software and decision support

FAA AAS Problems • Design contest IBM vs. Hughes – 1983 to 1988 –

FAA AAS Problems • Design contest IBM vs. Hughes – 1983 to 1988 – Detailed design of software but no coding – Voluminous documentation • Implementation by IBM – Many levels of bureaucratic oversight by FAA – Changing requirements, but not what controllers wanted

FAA AAS Alternatives • Modern display system + tracking software at High Desert Tracon

FAA AAS Alternatives • Modern display system + tracking software at High Desert Tracon – Needed to track fast military aircraft – Separate project funded by Do. D – Open system, COTS equipment (Sun), evolutionary progress of components • PC-based display developed in secret by FAA engineers

FBI VCF Federal Bureau of Investigations Virtual Case file • 9/11 – FBI failed

FBI VCF Federal Bureau of Investigations Virtual Case file • 9/11 – FBI failed to “connect the dots” and obtain a full picture – Agents could not correlate data from different sources and investigations – Work processes are paper-based • Need to replace obsolete Automated Case Support system – And consolidate many other databases and applications • Actually started in 2000, accelerated after 9/11 • Cancelled in Apr 2005 after costing $170 M

FBI VCF Plan • Replace essentially all diverse FBI IT infrastructure • 800 -page

FBI VCF Plan • Replace essentially all diverse FBI IT infrastructure • 800 -page requirement document – Specifies design details like web-page layouts rather than what the system should do • Aggressive deadlines but no milestones or detailed schedule • Use a “flash cutover” – log off old system on Friday and into new system on Monday

FBI VCF Problems • Lack of overall encompassing architecture • Many change requests once

FBI VCF Problems • Lack of overall encompassing architecture • Many change requests once agents saw prototypes (using spiral model) • Resulting schedule slips

FBI VCF Replacement: Sentinel • 6 years development, ~$440 M • In operation since

FBI VCF Replacement: Sentinel • 6 years development, ~$440 M • In operation since July 2012 • Success credited to shift from waterfall to agile in 2010 • Also use significant off-the-shelf products

USCIS ELIS US Citizenship and Immigration Services Electronic Immigration System • 8 M applications

USCIS ELIS US Citizenship and Immigration Services Electronic Immigration System • 8 M applications a year, all paper, requiring sending forms between offices • Plan: computerize 95 forms and pay 40 fees online by 2013, cost of 500 M$ • 2015 status: 1 form and 1 fee are working, cost 1. 7 B$ – Application to renew or replace lost green card – 2 other forms were pulled due to problems • Projected completion 2019, cost 3. 5 B$

USCIS ELIS Problems • 2008 contract to IBM using waterfall • Requirements complete only

USCIS ELIS Problems • 2008 contract to IBM using waterfall • Requirements complete only in 2011 • Takes up to 150 clicks to review an application – Navigate menus, open documents, etc. – Test: complete 0. 86 cases/hour, compared with 2. 16 with paper forms • Integration of 29 commercial software products • 2013: switch to multiple vendors, more agile – But 4 -month work cycle • 2015: switch to cloud platform