Case Study 4: Boeing's Virtual Fence

On January 14, 2011, Secretary of Homeland Security Janet Napolitano made it official:The Virtual Fence Project was to be officially cancelled. In her statement explaining the decision, Napolitano cited the difficulty in creating a unified, fully integrated security systemandpromised to "pursue a new path forward."What was left unsaid were the reasons that led to the final decisionprincipally, struggling with a too- complicated technical systemthat did not work but was leading to ballooning costs.Illegal crossing into the United States along the Mexican border has reached epidemic proportions in recent years.

Fear of drug smuggling,illegal aliens, andpossible terrorist incursions have made the issue ofhomeland security one of the major "hot buttons" in the political arena, both in Washington, DC, and within states located along the southern border as well as those in proximity to Canada. The problem is compounded by the sheer sizes of the borders involved. The Mexican/ U.S. border runs for nearly 2,000 miles, much of it across desert wastelands and inhospitable and remote areas.Establishing any sort of border security, in the wake of the 9/11 attacks, is a national necessity but a daunting and difficult task. The Department of Homeland Security (DHS), organized following the attacks on the World Trade Center towers, is charged with the responsibility of securing all borders and points of illegal entry into the United States, in cooperation with Customs and Border Protection.

As part of its mandate, it has developedplans for creating a more secure and stable borderwith Mexico toprevent the continuous flow of undocumented immigrants, drugs, and potential terrorists. For the first stage in this process, DHS proposed a project to physically and electronically seal the stretch of the desert between the United States and Mexico under a multibillion-dollar contract named the Secure Border Initiative Net (SBInet). President Bush in May 2006 called SBInet"the most technologically advanced border security initiative in American history."

A 28-mile stretch of desert, centered on Nogales, Texas, was to be the pilot stage in a project that eventually would be used to monitor and control some 6,000 miles of border with both Mexico and Canada. In late 2006, Boeing was selected as the major contractor for the SBInet project. Although better known for their military weapon systems, Boeing's Integrated Defense Systems Unit was made responsible for overall coordination of a massive system of towers as well as listening devices, motion sensors, cameras, and radar to be used to detect and help apprehend illegals crossing the border.

In fact, the U.S. government chose to outsource the entire project to private firms; that is, they expected that contractors would design the program's elements, build them, and then handle full oversight of their own work. In a nutshell, the system used a chain of 100-foottall towers that each scanned a 360-degree radius for a distance of 10 miles. Ground radar sensors also attempted to detect footsteps, bicycles, and vehicles. The first $20 million pilot phase, named Project 28 after the length of the part of the desert that it was supposed to cover, was to be completed by mid-June 2007. Boeing selected more than 100 subcontractors to build various components of the system, with its project managers maintaining overall control of the development process.

Unfortunately, their structure was unwieldy, and the project was further compromised by the sheer number of distinct elements and technical systems Boeing was attempting to integrate. The technical challenge of integrating systems including watch towers, sensors, radar, and specialized cameras was beyond anything Boeing had attempted before. The problem was particularly noteworthy when we consider that integration, in many ways, was the project. The various technical elements were difficult but attainable.

The challenge for SBInet lay in the ability of Boeing to find a means to bring all these new and unproven technologies together under one umbrella. So complicated was the challenge, in fact, that the virtual fence failed a series of initial tests, significantly delaying the full deployment of Project 28. Unfortunately, these technical and coordination problems were never resolved. In the nearly three years after original testing was done on one section of the fence, SBInet had cost the government $672 million dollars, with the end nowhere in sight. Although the total project cost was anticipated at $1.1 billion, congressional watchdog groups argued that the final cost of the project could soar to over $30 billion. Costs, in fact, were a sore point with the project from the time it was bid. Originally promising to complete SBInet for $1.1 billion, Boeing's revised estimates went to $2.5 billion and then, just a few months later, to $8 billion.

This rapid escalation of projected costs finally prompted a congressional oversight committee hearing, in which Boeing endured withering criticism from Representatives who questioned their motives in asking for more money and time to complete the project. In the meantime, beset by continuing problems, Boeing had also revised its estimates for the completion date to 2016, more than seven years after the date in the original plan.

A major concern was Boeing's pyramid-like management structure that critics said caused confusion and a lack of clear responsibility. Worse, it made it easier for hidden costs to be charged to the project. Because Boeing embedded multiple subcontracting layers in the Virtual Fence development, they were able to add charges at each level. The larger problem was the clear conflict of interest that emerged by placing Boeing in charge of project oversight, while allowing them to manage sub-contractors, and monitor the progress of the project. Not surprisingly, with this configuration, little information came to light about cost overruns or schedule slippages until quality and overrun problems were simply too large to ignore . . . or hide.

Critics compared this attitude of easy oversight and loose control to the huge problems that had plagued Boston's "Big Dig" construction project (see Case Study 8.2 in text). Admittedly, the problems that sank the SBInet project were complicated and came from multiple sources. Besides the technical challenges of managing 100 subcontractors, all required to provide critical components that Boeing would integrate, the project had effectively shut out most federal agencies and oversight groups. It was difficult to get accurate project status information given the government's decision to "farm out" border security to private contractors.

As a result, congressional investigators found that Homeland Security officials were simply standing by while Boeing provided information that was "replete with unexplained anomalies, thus rendering the data unfit for effective contractor management and oversight." Furthermore, many critics questioned the feasibility of the original intent of the project itself, wondering about the likelihood of ever effectively sealing a border that runs through some of the most inhospitable terrain in North America. Whether through a combination of poor oversight, over- optimistic scope expectations, or simple inability to make this cutting-edge technology work, SBInet remains an example of a significant program failure at the taxpayer's expense.

Questions

1. What problems do you see emerging from a project such as SBInet where the government allows the contractor to determine scope, manage all contractor relations, and decide how to share project status information with oversight bodies?

2. Consider the following two arguments: "The failure of SBInet is due to poor scope management," versus "SBInet failed because of poor oversight and project controls." Take one side or the other in this argument, and justify your response.