The Transportation Security Administration does not monitor two of its key performance requirements for its primary on-person screening systems at airport security checkpoints and therefore can’t be sure the body scanners are working as intended, the Department of Homeland Security’s top watchdog says in a new report.

The two operational requirements that the Advanced Imaging Technology (AIT) that TSA isn’t monitoring for are probability of detection rate and passenger throughput, Joseph Cuffari, the DHS Inspector General (IG), says in the report that is dated May 8 and was released on Thursday. The AIT systems were declared fully operational in September 2017, the report says.

The AIT systems are supplied by Leidos [LDOS], which acquired them from L3Harris Technologies [LHX] as part of $1 billion acquisition earlier this month of the security detection and automation business. The IG report says that TSA has purchased 962 AIT systems that are deployed to 340 U.S. airports to scan individuals for non-metallic threats hidden beneath their clothing.

TSA has four key requirements for its AIT systems to be fully operational, the report says. The requirements are probability of detection, which is classified but refers to the scanners identifying the location of a concealed threat on passenger, a throughput rate of at least 150 people per hour, a system availability rate of 98.5 percent, and safety of use metrics.

The IG says that TSA hasn’t been monitoring the performance of the two key requirements because the agency doesn’t have comprehensive guidance in place to do so. This could create blind spots for TSA, the report says.

“Without continuous monitoring and oversight, TSA cannot ensure AIT is meeting critical system performance requirements, a consistent weakness found in prior DHS OIG reports,” says the new report, TSA Needs to Improve Monitoring of the Deployed Advanced Imaging Technology System (OIG-20-33). As a result, TSA may fund and acquire future systems without knowing whether the technology addresses new and emerging threats, potentially placing the traveling public at risk.”

In 2015, the DHS Office of the IG reported that in 96 percent of instances Transportation Security Officers (TSOs) using AIT machines at eight airports failed to detect hidden threat items. Then TSA Administrator Peter Neffenger, who was relatively new to his job, said the system worked as designed when used properly. Neffenger instituted enhanced training for TSOs and also said at the time that the agency has prioritized passenger throughput at the expense of security.

In the latest IG report, TSA concurred with the watchdog’s two recommendations, which call for a strategy to monitor the probability of detection rates and passenger throughout of deployed AIT systems.

However, TSA did take issue with the IG’s assertion that the agency can’t ensure that equipment in the field meets performance requirements.

“TSA, in cooperation with DHS Science and Technology and industry, academia, and international partners, has demonstrated rigorous, disciplines, and agile requirements and capabilities development and acquisition and procurement processes,” Ryan Propis, TSA’s chief of staff, wrote in an April 1 letter to the IG that is included in an appendix to the report. “The research, development, testing, and procurement of transportation security equipment is intelligence-driven, risk-based, and focused on fielding the most effective security screening equipment against both current and emerging threats.”

In a separate statement yesterday, TSA it “concurs with both recommendations in the report, however, the report does not address the progress TSA is making to ensure security technology detection monitoring and performance in all security screening technologies, including AIT. TSA is actively developing procedures for monitoring detection requirements to improve in the areas addressed by the recommendations.”

Leidos has more than 2,000 of its ProVision AIT systems deployed at more than 250 airports and many non-aviation customers worldwide.