When US Department of Homeland Security (DHS) Secretary Janet Napolitano ordered the termination of the $1 billion virtual fence project called SBInet (Secure Border Initiative network) in January, she instructed US Customs and Border Protection (CBP) to develop "... a comprehensive border technology deployment plan that will build upon successful technology currently deployed and provide the optimum mix of proven surveillance technologies."
This "new way forward," DHS said at the time, "is expected to cost less than $750 million and will cover the rest of the Arizona border - totaling 323 miles." SBInet only covers 53 miles of the border.
In January, I asked how many believed that the cost of this new Arizona Border Surveillance Technology program would be less than $750 million?
Well, a US Government Accountability Office (GAO) report published last week is starting to provide some insight into the answer.
According to the GAO:
"CBP’s 10-year life-cycle cost estimate for the [new Arizona Border Surveillance Technology] Plan of $1.5 billion is based on a rough order of magnitude analysis, and agency officials were unable to determine a level of confidence in their estimate ... Specifically, GAO’s review of the estimate concluded that the estimate reflected substantial features of best practices, being both comprehensive and accurate, but it did not sufficiently meet other characteristics of a high-quality cost estimate, such as credibility, because it did not identify a level of confidence or quantify the impact of risks [associated with the Plan] ..."
For instance, one of the risk impacts not quantified are those involving the 5 integrated fixed-tower (IFT) systems consisting collectively of 52 fixed radar-and-camera towers, which are estimated to cost some $570 million alone for "... design and development, equipment procurement, production and deployment, systems engineering and program management, and a national operations center."
CBP originally assumed that the "... IFT systems would be able to access existing commercial communication networks in target deployment areas. [But] CBP officials said that this assumption is no longer valid in all cases and additional communication relay equipment will likely be necessary."
However, CBP told the GAO that it has sufficient risk contingency money to cover these unexpected costs. Nevertheless, the GAO is a bit skeptical, since the CBP hasn't performed "... a risk and uncertainly analysis to quantify the impact [of the risk] on the cost estimate." In fact, from reading the GAO report, the CBP seems to assume it can comfortably manage all of its program risks even though it hasn't performed any detailed risk assessments on them.
Given this apparent risk arrogant mind set that also plagued the SBInet debacle (so much for lessons learned), the GAO states:
"... it will be difficult for CBP to provide reasonable assurance that its cost estimate is reliable and that its budget request for fiscal year 2012 [of $242 million] and beyond is realistic and sufficient."
The CBP's $1.5 billion rough order of magnitude estimate for the Arizona Border Surveillance Technology Plan is currently divided into approximately $750 million for its acquisition and $800 million for its operations and maintenance. $185 million is already being spent during the current fiscal year on the effort.
The GAO also claims that the "CBP has not documented the analysis justifying the specific types, quantities, and deployment locations of border surveillance technologies proposed in the Plan." However, CBP argues that it has indeed done sufficient analysis to justify its spending request; i.e., trust our judgment and management expertise, says the CBP.
Just like everyone was supposed to during SBInet. Fool me once ....
In addition, the GAO report states that the CBP hasn't yet defined what the measurable "mission benefits" are related to its current spending plans. Apparently CBP's approach is to spend money now, figure out what was achieved later, and then, no doubt, declare that the system fully meets its success criteria.
Interestingly, the GAO also found in terms of "... actual performance of SBInet in terms of interdiction was only slightly different than if the system had not been present in the areas where it is deployed." Not bad value for $1 billion spent, eh? I wonder how many border patrol officers you could hire over a ten year period to patrol those 53 miles of Arizona border now covered by SBInet.
So, what is the likelihood again that CBP will spend under $750 million for the acquisition of Son of SBInet as promised? I would say it is rapidly approaching zero, if it isn't there already.
Robert N. Charette is a Contributing Editor to IEEE Spectrum and an acknowledged international authority on information technology and systems risk management. A self-described “risk ecologist,” he is interested in the intersections of business, political, technological, and societal risks. Charette is an award-winning author of multiple books and numerous articles on the subjects of risk management, project and program management, innovation, and entrepreneurship. A Life Senior Member of the IEEE, Charette was a recipient of the IEEE Computer Society’s Golden Core Award in 2008.