The unfortunate reality is that – for the safety of the people attending – security measures are now an absolute requirement in large public spaces, including performance venues, airports, train stations and even malls. While government, community, and public safety agencies pursue different approaches aiming to decrease the likelihood and frequency of potentially violent incidents, scientists are developing advanced technology solutions to help prevent them from occurring at the physical location.
Security in Motion
At the heart of a preventative approach to public-venue protection is the development of “seamless security” using a combination of an optical camera, which includes a depth sensor, and millimeter-wave radar. Currently, traditional airport screening uses a form of this radar, but in a static fashion. A person stands still, raises their arms, and the system scans them. This method is effective; however, it can cause bottlenecks and delays when there are lots of people trying to get through the same entry.
Seamless security technology tries to accelerate this process by scanning individuals while they are in motion.
With a seamless-security approach, the camera tracks an individual’s movement as they are scanned by the millimiter-wave radar. By combining the tracking and the radar data, the system provides better resolution and imaging quality without disrupting the crowd flow. To expedite screening, the security scanning data can also be fed to an integrated monitoring system with overhead video surveilance technology. In high-volume traffic environments, this real-time data flow is a significant benefit. It can identify objects of concern as well as unusual behavior, thereby prompting any preventative action.
While there are privacy concerns about the use of security scanning, the system can be configured to not reveal privacy-sensitive data to the operator. As common in current airport systems, the operator only sees an indication of where the suspicious objects are on the person body and possibly an image of the suspicious object. The system does not provide images of people scans to the operators. The purpose is to screen for suspicious objects, notably forms like guns or knives, and not invade privacy. It is a no-touch system, unless additional action is warranted.
Applications for seamless security include:
- Avoiding long lines and crowds at train or subway stations by scanning people while they are on the move.
- Conducting more efficient screening at large public events (e.g., large venues, parks, stadiums).
- Using high-precision screening and identification at business entrances to verify the identity of those entering the location.
- Establishing unique security zones in a building to identify for additional screening.
- Monitoring people on escalators as they move between floors in a facility.
- Improving screening at mall entrances.
- Combining camera and radar-systems imaging to identify suspicious items as well as anomalous behavior.
Waves of Innovation Ahead
As development of seamless security progresses, the likely first iteration, within the next three to five years, will be for more predictable, semi-static environments, such as people moving along a corridor or escalator or less precise scanning of people in motion. Performing millimeter-wave precision imaging for people in free motion requires the combination of optical cameras and radar data, a technology under active development. This combination requires algorithms that track the motion from the optical data and use the tracking to combine the radar data in a coherent way. The outcome is an accurate radar image, as if the person is scanned is a traditional static airport scanner.
To achieve this outcome, a security system requires four pillars—each containing an optical and a radar system—that coordinate to acquire a 360-degree screen of an individual. They can be placed along regular paths to provide non-invasive screening as people move along a building entrance, a stadium, an escalator or a business building corridor, for example.
A positive factor driving this development is that the costs of components needed for millimeter wave radar systems have come down, due to innovations predominantly in the automotive sector. A “system on a chip” pricing has made it more affordable to innovate new methods of large-venue security screening.
Artificial intelligence (AI) will continue to play a role in the future of seamless security. AI can improve the image models through machine learning, leading to better scanning and more accurate imaging. In addition, AI systems can automatically recognize suspicious objects, such as guns, knives or explosive devices, and distinguish them from safe items (e.g., keys, smartphone). To address privacy concerns, AI can help mask images that are not relevant to security scanning.
AI will also detect anomalous behavior in an overhead surveillance system and prompt further action form the security personnel. It will also be used for general security practices such as facial recognition to match a person’s identity with an individual wanting to gain access to a building. The use of AI should not remove the involvement of a human to qualify the accuracy of behavior and identity issues, especially considering concerns over AI safety, possibility of AI hallucinations, and incidents of deep-fake fraud.
Right now, seamless security will be achieved using different sensing modalities like optical sensors and radar. Longer term, terahertz imaging also looks quite promising, as it can also penetrate clothing while providing more precise imaging.
Ideally, seamless security in five or more years will combine biometric scanning for identity recognition and camera and millimeter wave screening for security scanning, giving people traveling in public spaces a more secure environment. As our security and identification needs evolve, technology follows to support them with improved hardware, algorithms and artificial intelligence. The goal should always be towards less intrusive yet more effective security systems.
