Machine Learning to Improve Access Control: Part 1

With the evolving dynamics of cloud storage and the ability to harness and proactively employ an ever-increasing pool of big data, AI in the form of machine learning and deep learning has become a disruptive technological force in the physical security industry. Advanced AI and low-cost network resources have significantly impacted video surveillance, which has been among the biggest beneficiaries of faster processing and impactful analytics. Building automation, fire systems, intrusion detection, and physical and network access control are all starting to incorporate AI functionality.

AI is increasingly taking a role in making exterior and interior entrances more secure.This technology is moving towards improving system functionalities, including: distinguishing people from objects at a facility perimeter and interior entrances; deterring piggybacking; identifying and analyzing potentially lethal objects and dangerous people; and helping to define secure areas in and around buildings creating a more defensive risk posture for a location.


As technology continues to converge and the edges of different traditionally separate technologies start to merge, we face the challenge of how AI may practically support entry solutions such as security revolving doors, turnstiles, and swing doors. A disconnect between the objectives of the building owner and building code regulations can further complicate the security blueprint. But with 5G communications on the horizon, at least for those in major metro areas, multiple systems will be able to communicate seamlessly and instantly. With billions of connected IP devices generating data, Machine Learning systems will have an enormous amount of data to run through algorithms and improve performance in the budding field of intelligent access control.

Because legacy security entrances do not have AI built into their technology, integrating intelligence into secured entrances requires a collaborative effort with a third party solutions provider. Video analytics are increasingly deployed to address use cases such as people detection, piggybacking, dangerous object detection and facial recognition among other issues relevant to secured entrances. The increased integration of AI providers with traditional security entrance partners has resulted in improvements, such as price, speed, ease of use and usability. It also includes the use of machine learning to improve algorithms over traditional modeling and correlation approaches, and integration with other systems and sensors.

The Plan Moving Forward
Security entrances and mantrap portals often combine a number of systems, sensors and requirements. Portals by their nature are an integrated solution combining access control, video surveillance, mechanical hardware, sensors and design. These systems are a micro-scale example of what we can expect to see rolling out in the next decade. As devices move to wireless and 5G increases the throughput and number of IP devices, trillions of data points will be created for algorithms to process.

The limiting factor at that point will be only processing power and the limitations of human ingenuity in programming. Nonetheless, the access control, video surveillance, alarm, and community management/smart home technologies will begin to bleed into one-another and be able to intelligently share relevant information to improve performance, all with almost zero latency. The implications, should this play out ideally, include hybrid systems capable of operating nearly without human intervention, a reduction in the need for monitoring center services, and better response and accuracy for access control and security systems alike.