The X-ray machine has been a staple component of aviation security operations for decades, but as throughput rates increase as fast as the number and types of threats we face, technology has had a tough time keeping up. Joseph Paresi discusses the benefits of implementing computed tomography (CT) at checkpoints and examines the progress being made in the technology.
At a recent security conference, an airport security director claimed that it would be a long time before he would deploy computed tomography (CT) at the checkpoint. When asked why, he said that he “could not afford to have a system that is slower than existing X-ray lanes”. If that were true, I would agree with him, but the real facts associated with checkpoint CT throughput are that they are faster…when used properly.
It first needs to be noted that the need for CT technology at the checkpoint is 100% mandatory and, like with the hold baggage mandate, every aviation checkpoint in the world will be using CT systems soon. Terrorists have studied X-ray systems and know how they can be beaten. Airports and airlines cannot wait and react. They must be proactive before the incident!
“…terrorists are no longer using metal detonators, formerly a key indicator of a potential improvised explosive device…”
The Office of the Inspector General (OIG) recently presented the results of an undercover test performed in 2017, which were consistent with the extremely poor results of the 2015 undercover test in which agents missed 96% of hidden threats. Immediately after the OIG briefing, Congress met with the TSA Administrator, Admiral David Pekoske. At that meeting, Representative Michael McCaul, chairman of the Congressional Homeland Security Committee, urged the Administrator to, “Deploy CT scanners as quickly as possible to all U.S. airports and the ten airports considered the highest threats that send flights to the U.S.” Afterwards, Representative Bennie Thompson stated, “Following the hearing, I remain convinced that DHS must act — without delay — to acquire and deploy updated screening technology, including Computed Tomography, at our nation’s airports. Such technology has proven more effective at detecting threats than existing systems and is currently deployed successfully elsewhere in the world.”
We need to recognise that poor undercover test results were not due to the fault of the screening agents. They were the result of the inherent weaknesses of the X-ray technology. No one can effectively review thousands of two-dimensional X-ray images and visually identify an explosive threat. This is punctuated by the fact that terrorists are no longer using metal detonators, formerly a key indicator of a potential improvised explosive device (IED). Agents previously looked for a blue coloured image of a metal detonator within the orange coloured section of the bag image (as orange represents organic material such as explosives). Without a metal detonator present, there is no clear way to ascertain if an object is a threat or a benign object. The industry understands this fact and every major security supplier has developed a checkpoint CT offering in response. Simply put, the classified threat list, as defined in the new ECAC CTP and TSA APSS Certification standards, can only be properly addressed today with CT technology. Regulators are pushing for these detection solutions and many international airports are examining the use of checkpoint CT technology. In the U.S., the TSA Administrator and his staff have stated their intention of deploying CT technology at 2400 U.S. Airport checkpoints. Bravo, Admiral!
Another consideration are today’s operational costs. At one major airport, the cost of implementing the additional TSA security measures (i.e., interviews, extra screening, etc.) is estimated to be costing $30 million per year or $500 per passenger! TSA has agreed to eliminate the need for these extra security measures with the use of CT at the checkpoint.
“…TSA is planning for automated restricted items detection…”
At present, all checkpoint CT systems operate with a belt speed of 15 centimetres per second or greater. This means that for an average-sized, standard carry-on bag of 511 millimetres in length with spacing, CTs could scan over 900 bags per hour. Long trays cut this throughput to around 700 trays per hour, but there is more to consider: bags or trays lined up back-to-back will result in lead safety curtains being open for too long at this belt speed. To solve this, vendors have had to either add spacing between bags, reducing the system throughput, or develop an alternate approach to address safety compliance. However, a throughput of 600 trays per hour can be achieved within radiation safety compliance. This throughput is far greater than that of an X-ray scanner, which typically maxes out at around 375 images per hour or 150 passengers per hour using a C1 (removal of electronics and liquids) concept of operation (ConOps).
In operation, using C2 ConOps (leaving the electronics in the bag), we are actually achieving a throughput of over 300 trays per hour or 170 passengers per hour (C2 reduces the number of bins per passenger from around 2.5 to 1.8 each). In addition, around 80% of the restricted items identified are bottles. When the jump is made to C3 ConOps (leaving electronics and liquids in the bag), throughput dramatically increases to over 450 trays per hour or up to 300 passengers per hour (C3 reduces the trays per passenger from 1.8 to 1.5). However, this can go even higher; the key factor affecting throughput is agent review time.
As X-ray scanners provide only two-dimensional images, there are fewer views for an agent to review. CT scanners allow agents to view both two- and three-dimensional images, with many more features such as slice, slab and virtual laptop removal mode. The visual detection performance can therefore be much better but sometimes, better is the enemy of good enough.
If an agent took only six seconds on average to review images, the 600 images per hour would be realised. An airport could multiplex the images and use more than one agent per scanner to achieve a 6-second goal, but that is a cost trade-off not everyone is willing to accept. Fortunately, there are other ways to address this target.
Originally, suppliers provided their checkpoint CT scanners with limited colouring. Automatically detected threats were coloured red while high-density objects, representing restricted items including guns and knives/sharps, were coloured in blue. Using this ConOps, two independent tests were performed to measure CT effectiveness in identifying restricted items – one by a Heathrow/Manchester airport team and another by agents in Amsterdam. In both cases, the agents achieved 98.5% accuracy in the visual detection of restricted items – way better than has ever been achieved before. Then the airports asked for TriMat coloured images, like those of an X-ray scanner, in addition to a two- and three-dimensional image display. This ConOps resulted in agent review times of between 10 and 20 seconds. However, if the CT scanner can automatically detect the threats, what value is there to TriMat colouring except for special application? The best and fastest way to use CT at the checkpoint is by having the agent focus only on the identification of restricted items, which only requires the blue image feature (Aside: TSA is planning for automated restricted items detection in future software releases). In this mode operator review times have been measured as low as 6-7 seconds per image, which provides excellent throughput and increased passenger flow. If operators are consistently achieving 98.5% accuracy on restricted items detection, TriMat images cannot realistically improve performance.
The final factor in throughput is the number of bags that require secondary hand search. The contributing factors are the false alarm rate and the number of restricted items identified by the agent for hand search. Using a target of 450 BPH, a 15% alarm rate will result in 68 bags per hour that require hand searches. If each bag search takes 40 seconds, the total time spent on bag searches in an hour would be 2720 seconds, or about 75% of the time. This can be improved if the secondary search agents use on-screen alarm resolution to clear some of the alarms, which could be performed in less than half of the 40-second hand search time. Also, the fact that the images shows the exact 3D location of the potential threat, should reduce the search time, possibly in half. At a rate of 600 images per hour throughput with a 15% total alarm rate, 3600 seconds or one hour would be required in an efficient operation. As could be the case with image analysis, more than one agent can also be reallocated to perform hand searches.
As it stands, current threats are driving the need for checkpoint CT systems. The recommendations presented above offer a balanced approach with greatly enhanced security. Increased throughput, the need for fewer scanners, as well as room for passenger traffic growth, without the need to expand the checkpoint area, are all available under CT implementation. All things considered, checkpoint CT technology will provide both better security and a much better passenger experience.
Joe Paresi is the founder, Chairman and Chief Executive Officer of Integrated Defense and Security Solutions, IDSS, which has developed the DETECT™ 1000. Prior to IDSS, he was the co-founder and Executive Vice President of L-1 Identity Solutions, Inc. Prior to L-1, Joe Paresi served as Corporate Vice President of Product Development at L-3 Communication Corporation and President and founder of L-3 Security & Detection Systems, where he led the development and deployment of the L-3 eXaminer 3DX 6000 TSA Certified Explosive Detection System and the ProVision millimetre wave body scanners. Mr. Paresi also served as Corporate Director of Technology for Lockheed Martin and Loral Corporations.