The Aviation Subcommittee will meet on Wednesday, March 14, 2001, at 1:30 p.m. in Room 2167, Rayburn House Office Building to receive testimony from the Federal Aviation Administration (FAA), the Department of Transportation's Inspector General (IG), and other interested parties on FAA’s efforts to modernize the Air Traffic Control system with a focus on the Standard Terminal Automation Replacement System (STARS).

STARS is a joint FAA and Department of Defense (DoD) procurement program intended to replace the aging Automated Radar Terminal System (ARTS) at 173 FAA TRACONs, 15 support sites and up to 199 DoD terminal facilities. STARS will work in conjunction with digital radar systems to allow air traffic controllers to track aircraft within the terminal area. The new equipment and software will be based on a digital platform and provide higher-resolution screens with color capabilities and higher system reliability. STARS can also be expanded to meet increased traffic demands and accommodate new automation functions.

The FAA air traffic control modernization program has a history of cost overruns, delays and failures; STARS has been no exception. STARS is more than 2 years past the original date for operational readiness and is still several years away from becoming operational.

The Committee has been concerned with the management of the STARS program since STARS was a component of FAA’s failed Advanced Automation System project (AAS). The Aviation Subcommittee held several hearings on AAS and, on March 5, 1998, held a hearing specifically on STARS.

BACKGROUND
In 1983, FAA announced plans to modernize its air traffic control system, which is a nationwide network of surveillance equipment, navigation aids, computers, communication systems, and personnel whose mission is to safely guide aircraft throughout our air traffic system.

FAA's major modernization project, the Advanced Automation System (AAS), was originally estimated to cost $2.5 billion with a completion date of 1996. The program, however, experienced numerous delays and cost overruns, which were blamed on both FAA and the primary contractor, IBM. In 1994, FAA cancelled part of the program and split the remaining systems into three phases, and in several cases, re-bid the contracts. In September 1996, FAA recreated the portion of the AAS program known as the Terminal Advanced Automation System or TAAS (this time calling it STARS), and awarded the contract to Raytheon.

According to the General Accounting Office (GAO), almost $1.5 billion of the $2.6 billion spent on AAS was completely wasted. Only $1.1 billion worth of the work done on AAS was reused in the successor programs. The IG estimates that only $20.9 million of the $396.6 million spent on TAAS was salvaged for STARS. In addition, there were approximately 100 FAA employees working on the AAS project from the mid 1980s to 1994; GAO did not estimate the cost of the FAA employees. GAO also estimates that FAA has spent an additional $660 million on interim projects to sustain and enhance current equipment.

The 1996 contract focused on acquiring STARS through commercial or "off-the-shelf technology." FAA believed that this approach would reduce costs and limit risk since several companies had systems in place in foreign countries. However, issues arose that forced FAA and Raytheon to restructure the original contract and to engage in significant amounts of software development.

One issue that contributed to much of the project delay and cost growth in the STARS program was the lack of coordination between the FAA and air traffic controllers and service technicians during the early stages of development. This led to a STARS design that could significantly impact the controllers’ ability to control aircraft.

In December 1997, the National Air Traffic Controllers Association (NATCA), the Professional Airways Systems Specialists (PASS), FAA, and Raytheon formed a working team to deal with the problems. The team identified 98 human factor issues and developed a package of enhancements (referred to as Computer Human Interface enhancements or CHI) to address the problem.

FAA and Raytheon restructured the contract to develop STARS in three major phases: Early Deployment Configuration (EDC); Full Service 1 (FS1); and Full Service 2 (FS2). EDC would provide controllers with new digital workstations but would continue to operate on the existing ARTS system. FS1 would provide full STARS functionality with only 80 percent of the FAA CHI enhancements. FS2 would provide full STARS functionality with all CHI enhancements.

STARS EDC has been operational at El Paso and Syracuse over a year and a version of full STARS began operation at Elgin Air Force base in June 2000. FAA expects to deploy STARS FS1 at El Paso in April 2001.

FAA contends that all of the major development issues have been addressed and that STARS is on track to proceed with deployment. However, as much as 16 percent of the STARS software has not been fully tested. FAA believes it has mitigated much of the risk through the fielding of EDC and the initial testing of software.

STARS FS2 will undergo tests and evaluation this summer. If no major problems are discovered, the first FS2 will be scheduled for initial operating capability (IOC) in 2003.

ISSUES
Development of a Contingency Plan
Despite risk mitigation efforts, there is concern over FAA’s approach to risk management. While EDC has been successful, it operates on the old ARTS platform, while Full STARS will be based completely on the STARS platform. Some feel that FAA needs to develop a contingency plan in case STARS cannot be deployed as scheduled. The ARTS IIIE system could be a potential back-up system.

ARTS IIIE is an upgrade of the older ARTS IIIA system. In 1997, when it was apparent that STARS was several years away from operation, FAA contracted with Lockheed to refresh the older technology. Five ARTS IIIEs were deployed to the busiest TRACONS. ARTS IIIE is similar in functionality to STARS and has been operational since 1998.

Deployment Schedule
Currently, some of the most significant cost and schedule risks to the STARS program are associated with the deployment of STARS FS2. Last week, FAA unveiled a schedule that would have the first FS2 site in Oregon by 2003 and that would have complete implementation of STARS by 2008. This means that, on average, FAA will have to commission one site a week to meet its schedule. Some consider this schedule to be too aggressive.

FAA will begin by replacing the oldest system first, the ARTS IIIA. Over the years, each ARTS IIIA location has developed unique software for site-specific functions (known as "patches"). FAA has completed audits of 3 of the first 35 proposed sites and has found 23 patches. There is some concern that duplicating these patches could require substantial software development and strenuous testing. FAA is in the process of evaluating each patch.

Many of the facilities also must receive upgrades to coincide with STARS installation. Items such as leaky roofs, improved power systems, communication lines, and heating and cooling issues will have to be addressed before any STARS equipment can be installed. FAA has been cataloging facility requirements and will schedule improvements to be completed before installation.

Escalating Cost
STARS has a history of escalating costs. The original STARS program (TAAS) was estimated to cost $460 million. In 1993, after an internal review of all modernization programs, FAA revised its estimate of TAAS to $654 million. In 1996 when the contract was awarded to Raytheon, the estimated cost of the STARS programs was $940.2 million. In 1998, FAA and Raytheon restructured the contract to deploy STARS in various phases and to address human factor issues. This increased the cost estimate to $1.4 billion ($273 million for human factor changes).

The FAA is currently developing an estimate for the additional costs to cover increased personnel and process improvements such as rover teams, a portable TRACON, mobile labs and training facilities. The IG estimates that these changes could cost an additional $300 million.

ASR-11 and Digitizers
One of the main advantages of STARS is its digital platform. This allows the system to effectively integrate large amounts of data quickly and accurately. STARS requires digital feeds from the radar systems to track air traffic and weather systems. Most radar systems in use at FAA TRACONS (ASR-7 and ASR-8) provide analog feeds. The ASR-9, the only digital radar currently in the FAA inventory, is only in place at 57 TRACONs.

Timely deployment of STARS depends on the next generation radar, the ASR-11, currently under development. In recent months, the ASR-11 program has experienced setbacks that will delay its delivery.

Some have suggested that if the ASR-11 is not available when FAA has completed STARS installation at all TRACONs with ASR-9s, the FAA has the option of installing digitizers to bridge the gap between STARS and older analog radar. However, FAA has raised some concerns over the cost and effectiveness of digitizers. Digitizers would only provide a temporary fix at a cost up to $1 million per installation. Digitizers have also had problems distinguishing between weather, topography, obstructions, birds and aircraft. A possible solution would be to install digitizers on all radar modes, potentially doubling the cost.

Radar Synchronization
Unlike the ARTS system, STARS will be able to use feeds from multiple types of radar systems. This system, known as a Mosaic system, allows controllers to track targets over larger areas and provides a built-in backup system.

While a Mosaic has many benefits, it does create technical issues. Short-range terminal radar such as the ASR-9 and the ASR-11 rotate at a rate of once every 4.6 seconds. Long-range en route center radar systems such as the ASR-4 rotate at a rate of once every 12 seconds. This lack of synchronization can cause ghost images and target jumping. FAA is working to address these issues.

Security
STARS is designed for remote monitoring. This allows a single facility to monitor the performance of several STARS systems. Remote access may create added security risks from computer hackers. This issue is still under discussion.
 
 
 

STATEMENT OF STEVEN ZAIDMAN, ASSOCIATE ADMINISTRATOR FOR RESEARCH AND ACQUISITIONS FEDERAL AVIATION ADMINISTRATION, BEFORE THE COMMITTEE ON TRANSPORTATION AND INFRASTURCTURE ON THE STATUS OF STARS, MARCH 14, 2001

Chairman Mica, Congressman Lipinski, Members of the Subcommittee, I am pleased to have this opportunity to speak to you about the Federal Aviation Administration’s effort to develop and deploy the Standard Terminal Automation Replacement System (STARS). I am especially pleased to be the first FAA witness to testify before the new Chairman of the Subcommittee, Chairman Mica. On behalf of the FAA I would like to say that we considered our working relationship with the former Chairman of the Subcommittee, Chairman Duncan, to have been a substantive and productive one, and we look forward to continuing in that vein with Chairman Mica and the other Members of this Subcommittee.

When completed, the installation of STARS will have replaced the display screens and the computer automation systems for 173 FAA terminal radar approach control facilities (TRACONS) and 102 Department of Defense air traffic control facilities. TRACONS generally control air traffic that is within 40 to 60 miles of an airport from which the aircraft departed or is to arrive. Some of the equipment and software STARS will replace is more than 20 years old, particularly at our 54 mid-to-high level demand TRACONS, which include Baltimore-Washington, Dulles, Philadelphia, and Orlando. These older systems can no longer accommodate any additional demand in capacity, which is projected to grow at a rate of 10% every four years. STARS will increase the number of "tracks" (aircraft targets displayed) from 900 at these mid-level facilities to 7000. This program is therefore an important component in the agency’s overall modernization effort. STARS will also provide a platform for future enhancements to air traffic controllers in areas such as new free flight capabilities.

While we have been proceeding with the STARS program, and while we work towards its full national deployment, which is scheduled to be completed in 2008 for FAA facilities, we have taken other interim steps to help ensure that we continue to sustain our ability to meet current demand. Between 1998 and 1999, for example, we modernized the computer automation system at all five of our highest demand terminal facilities, and we provided new color displays to four of these facilities last year. We also modernized the computer automation system at 131 low-to-mid level demand TRACON facilities across the country between 1998 and 2000. We have not only been modernizing the terminal air traffic control infrastructure, but have also made great strides in modernizing our en route air traffic control systems by replacing the HOST computers and providing new color displays at all 20 of our en route air traffic control centers.

I understand and appreciate that Members of this Committee are concerned that the STARS program had suffered from delays and cost growth, so I would like to take a moment to review the history of the program so that you will better understand where we are, how we got here, and most importantly, where we are headed.

In 1994, the FAA pursued a commercial-off-the-shelf product for the modernization of terminal automation systems. After considering various off-the-shelf products, Raytheon was awarded the STARS contract in September 1996. At that time, following several years of development, Raytheon’s commercial product had been operational for about a year in Oslo, Norway, and at two facilities in Oman. Consequently, we believed STARS could be purchased, modified slightly, and deployed relatively quickly.

Not long after the STARS contract was awarded, it became clear that the modifications necessary to adapt the equipment to the demands of the National Airspace System (NAS) of this country were going to be far more complex than were originally anticipated. The air traffic control and maintenance work forces identified numerous human factors issues that they strongly believed, if not addressed, would have compromised the safety of air traffic. It is easy in hindsight to recognize that a system that was working in Norway and Oman would not necessarily transfer easily to the very different challenges facing U.S. airspace. One difference is the sheer volume of air traffic that is controlled at our busiest facilities is significantly greater than what is seen at many international facilities. A second difference is the nature of air traffic control, which is more complex in the United States than it is abroad. In the U.S. we control air traffic on a principle of "first come, first served," while many overseas systems utilize a more centrally planned "time slot" approach. Their approach allows the controller significantly more time to plan and control their assigned aircraft at a more steady and slower pace. Our approach is more demanding on the controller, but allows the users of the system enormous flexibility. Because of these differences, our air traffic controllers have developed several computer-human interfaces and tools over the years that have become a standard part of the automation systems and displays that they use. Many of these computer-human interface features were not present in the commercial product that Raytheon had developed for the international market.

For the next 2 years, FAA worked extensively with Raytheon and our work force to address the human factors issues identified to the satisfaction of all parties. The ever increasing demands being placed on the NAS in combination with the complexity of the airspace in certain terminal environments made moving to a system with significant configuration differences problematic. The controllers and system specialists would be expected to transition to this new system with no disruption to traffic and absolutely no compromise of safety. It became clear that in order to meet these two essential goals for transitioning to the new equipment, changes would have to be made that would make the switch as transparent as possible.

In October of 1999, we briefed Congressional staff, including the staff of this Committee, to inform them about the changes that were made to the STARS contract, the cost implications that resulted from those changes, and the timeframe in which final software development and system deployment would occur. Essentially, we converted the commercial-off-the-shelf contract to a contract that had significantly more development work than had been originally planned. Since the time of that briefing, the FAA has largely remained on schedule and has not suffered any additional costs.

The largest challenge we faced in the last two years was to complete the development of the new software to incorporate the computer-human interface changes identified by our work force. That challenge is now mostly behind us. By the end of this month, we will have completed the design and coding of 84% of the software that will be needed to start the national deployment of the full STARS product. The balance of this software design and coding will be completed by the end of September of this year. We are also confident that this software will meet our requirements, as we have had extensive user involvement in its design and validation to date. Formal system testing that will be done over the next 20 months for the various STARS incremental software elements will validate its operational effectiveness and suitability.

If the major risk of software development is largely behind us, the major challenge that lies ahead is efficiently deploying the system to 173 FAA facilities by 2008 and to 102 DOD facilities by 2011. Since last summer, we have taken steps to identify and manage the challenges associated with future deployments. We formed an agency-wide working group to identify and manage the risks, to develop streamlined deployment processes, and to revalidate the amount and type of resources that would be needed to do the job. The issues they investigated included how to provide the necessary training for our controllers and systems specialists, align the STARS budget and schedule with those of anticipated TRACON and tower building modifications/construction as well as the budget and schedule of the ASR-11 digital radar, and coordinate the identification of site specific needs. All site specific needs must be individually evaluated in order to determine which ones should be incorporated. Many of the recommendations of the working group are in the process of being implemented and we are continuing to investigate other issues and recommendations prior to deciding how to proceed. Managing all of the risks involved in the efficient deployment of the system will require the same level and degree of senior management attention that we have given to the software development over the last two years. Our schedule provides for 20 months to continue to prepare and to mitigate these risks prior to the start of national deployment.

However, I do not want to leave this Committee with the impression that we are waiting to complete full STARS development before we deploy needed upgrades. We have worked with Raytheon and our work force to develop an early STARS capability that provides the full STARS displays with some basic hardware and software. We call this system EDC, for early display capability. We have had two sites operational (El Paso and Syracuse) for the last 14 months. Between late FY 2001 and early FY 2003, we will deploy EDC at an additional 11 facilities.

Modernization is a continual progression. The Administrator has spoken many times about the FAA’s restructured approach to modernization; that it is evolutionary, not revolutionary; that we are building a little, testing a little, deploying a little. Since we restructured STARS in 1999, we have been generally pleased with the results of this approach, including our ability to detect and address problems earlier in the process. Through our continued efforts and good communication with the operators and maintainers of our system and with system users, we believe that STARS modernization will progress and provide a reliable and efficient system that will support the world’s most complex and safest airspace.

Thank you for the opportunity to testify before you today. Mr. Chairman, I will be happy to answer any questions at this time.