Why the perimeter is becoming the next battleground for retailers and commercial operators.

Organized Retail Theft (ORT) has evolved far beyond isolated shoplifting incidents. Today, the FBI and its law enforcement partners increasingly view ORT as organized criminal activity involving sophisticated theft networks that operate across multiple retailers, jurisdictions, and supply chains. Intelligence sharing, coordinated investigations, and multi-state enforcement initiatives reflect a growing recognition that organized retail crime has become a national security and economic issue rather than simply a retail loss prevention challenge.

That shift was confirmed in May 2025 when a nationwide Organized Retail Crime Blitz involving more than 100 law enforcement agencies and major retailers resulted in approximately 500 arrests across more than two dozen states. The initiative highlighted a growing emphasis on dismantling organized theft networks through collaboration between retailers, investigators, and prosecutors rather than simply arresting individual offenders.

Industry organizations have reached similar conclusions. The National Retail Federation, Retail Industry Leaders Association (RILA), and Loss Prevention Research Council (LPRC) have all identified organized retail crime as a growing threat that increasingly extends beyond traditional in-store theft and requires stronger collaboration between retailers, law enforcement, and security providers.

Organized Crime Is Moving Upstream

Recent prosecutions demonstrate both the sophistication and scale of today’s organized retail theft organizations.

One of the largest ORT investigations in recent years involved a Queens, New York-based theft ring accused of stealing more than $2.2 million in merchandise from 128 Home Depot stores across nine states. Authorities charged 13 defendants in a 780-count indictment after linking the organization to more than 300 theft incidents involving power tools, appliances, building materials, and other high-value merchandise.

On the West Coast, authorities dismantled what prosecutors described as the largest Home Depot theft ring in company history. The organization allegedly conducted more than 600 thefts across dozens of Southern California stores, generating approximately $10 million in stolen merchandise before investigators recovered millions of dollars in property and arrested multiple suspects.

California has significantly expanded its organized retail theft enforcement efforts through increased funding, specialized investigations, and closer collaboration between retailers and law enforcement. As a result, state and local agencies arrested more than 29,000 suspects, referred nearly 23,000 cases for prosecution, and recovered more than $226 million in stolen property over a two-year period. Recent enforcement efforts alone resulted in more than 3,300 arrests and nearly $36 million in recovered stolen goods between July and September 2025.

These investigations point to an important shift in criminal strategy. Organized theft groups increasingly look beyond the retail sales floor, identifying opportunities throughout the supply chain where valuable assets may be stored with fewer people, less surveillance, and larger opportunities for theft.

Mobile Security Trailers Move Into The Mainstream

There is a version of perimeter security that most organizations still rely on: a camera detects that a line has been crossed, an alarm reaches an operator, the operator reviews the video data, and a decision is made. On paper, that chain of events sounds reasonable. In practice, it’s cumbersome and labor-intensive.

The problem is not that these systems fail to detect things. It is that they detect too much of the wrong things and too little of the right ones. Security teams end up managing a constant stream of false alarms, and operators tasked with evaluating every notification across dozens of camera feeds cannot realistically keep pace with the volume. Then, when a genuine threat does emerge, the gap between detection and response is often where the damage occurs.

The good news is that recent AI advances are fundamentally changing this equation, not by replacing human judgment, but by helping make sure human attention goes where it is actually needed. The 2026 SIA Security Megatrends report identifies AI as enabling a historic shift from detection and response toward genuine prevention. For perimeter security, that shift is already underway, and agentic AI systems that don’t just detect but can also verify and act, are the next step in closing the loop.

The Problem with Simple Models

Camera-based analytics, the kind built into most smart cameras, operate with limited compute power. They run relatively simple machine learning models: line crossing, person detection, vehicle detection, etc. These models are fast and inexpensive, but they are not very precise. A fox running across a monitored path looks enough like a person to the camera’s model that an alarm fires. A person wearing white against a snowy background may not register at all. The camera does its best with what it has, but what it has is not enough to reliably separate a real threat from a false one.

The result is alarm overload. Not because the technology is broken, but because models running on the camera are optimized for speed, not accuracy. According to the mentioned SIA Megatrends report, AI-driven solutions are helping reduce false alarms by approximately 60%. That improvement doesn’t come from replacing camera-based detection. It comes from adding a more intelligent validation layer on top of it.

Cascading AI: Smarter by Design

An approach reshaping perimeter security today is called cascading analytics, and the logic is straightforward: use the least compute power necessary at each stage, and only escalate the analytics when there is actually something worth examining more closely.

In practice, with cascading analytics, a camera detects a line crossing and raises an alarm. Rather than sending that alarm directly to an operator, the system passes the video clip to a more powerful server-side analytics engine with the GPU horsepower to run significantly more sophisticated models. That second layer checks the camera’s work. If it confirms a person, the alarm proceeds. If it identifies a fox, a shadow, or a flag moving in the wind, it suppresses the alarm before it ever reaches anyone. Because the advanced processing only activates when the camera has already flagged something, the compute is spent exactly when and where it is needed, not continuously across every feed.

Cascading can also be extended to a third tier, such as with generative AI models capable of producing open-ended scene descriptions. When an event has been confirmed by the server-side layer, the clip can be escalated to a vision language model that generates a detailed text summary, not just “person detected” but a description of what that person was wearing, how they were moving, and what they were carrying. This level of contextual detail is enormously useful when operators need to act quickly or when the video data will be used as evidence.

Think of it this way. A cascading AI architecture is like having access to a highly experienced but expensive consultant you don’t want to call every five minutes. You wait until the situation genuinely warrants it, and then you get exactly the insight you need. The camera handles the volume, the server-side analytics handles the verification, and the generative AI handles the interpretation. Each layer does what it does best.

Behavioral pattern recognition can add another dimension to this. A vehicle appearing near a perimeter three evenings in a row, lingering just long enough to observe before moving on, would not trigger a traditional alarm on any single night. An AI-driven system that correlates activity across days of recorded data can flag that pattern as anomalous and surface it for review before an incident occurs. That’s not just detection, that’s prevention.

From Verified Alarm to Coordinated Action

Filtering and verifying alarms are meaningful improvements, but the real transformation comes from what happens next. Agentic AI refers to systems capable of interpreting data, making decisions, and initiating action without waiting for a human to read a notification and decide what to do. For perimeter security, this means the platform itself can trigger a coordinated response the moment a genuine threat is confirmed: lights activate in the area of intrusion, an automated PA announcement warns individuals in a restricted zone, the access control system locks down entry points, and the right team members receive an alarm with verified video pushed directly to their workstation or mobile devices.

This shift fundamentally changes how security operations function. AI-powered platforms increasingly handle labor-intensive tasks such as monitoring feeds, ranking alarms, and managing access decisions, which allows personnel to transition from reactive notification watchers to proactive risk analysts.

And when an incident occurs, integrated evidence management capabilities make it possible to clip, annotate, and share relevant video data directly within the platform, without the overhead of manual exports or separate chain-of-custody workflows. For organizations with compliance requirements or law enforcement relationships, that capability is not just a convenience; it’s an operational necessity.

Why Open Platform Architecture Makes All of This Work

None of what is described above functions well in a closed, single-vendor environment. Perimeter security is an inherently multi-system problem. Cameras, radar, access control, intercoms, lighting, PA systems, drones, and mobile notification tools are rarely all from the same manufacturer, and they should not have to be.

An open platform video management architecture provides the common integration layer that makes it all work, connecting disparate hardware and software through standardized APIs and manufacturer-agnostic design. Camera alarms flow into server-side validation. Confirmed events trigger downstream response systems. Generative AI summarization layers are used when the situation calls for it. A facility that starts with basic line-crossing detection can grow into a fully cascaded, AI-driven security environment without replacing its core infrastructure. New analytics capabilities can be added as they mature. The platform grows with the organization rather than constraining it.

The Responsible Path Forward

As AI takes on a more active role in security operations, it is worth clarifying what that means. Agentic AI should augment human decision-making, not circumvent it. The most effective deployments use AI to surface verified information, execute pre-approved response protocols, and ensure the right people have what they need to act quickly, while keeping humans in the loop for decisions that require judgment, context, and accountability.

Perimeter security has always been about protecting what matters most. The tools have evolved considerably, from fences and floodlights to thermal cameras and radar to cascading AI analytics and agentic response workflows. What has not changed is the fundamental goal: knowing what is happening at the boundary of your environment, verifying whether it poses a real threat, and responding before harm occurs.

The technology to do all of that, intelligently, efficiently, and at scale, is being activated today. The question now is how well organizations are positioned to put it to work.

By Barry Norton, Fellow, Milestone Systems

Why Hostile Vehicle Behavior—Not License Plates—Will Define the Next Generation of Perimeter Security

For more than two decades, License Plate Recognition (LPR) has served as a valuable tool for identifying vehicles entering or leaving a property. Whether deployed by law enforcement, parking operators, or corporate security teams, the technology has focused on answering a single question: What vehicle is this?

Increasingly, security teams need to know far more than a vehicle’s identity.

Today’s most significant threats rarely arrive unannounced. Organized retail crime crews conduct reconnaissance days or weeks before striking. Cargo theft organizations survey distribution centers before targeting shipments. Individuals intent on workplace violence often visit a facility multiple times before an attack. Activists, copper thieves, catalytic converter thieves, and trespassers frequently exhibit recognizable patterns of vehicle movement long before a crime occurs.

In many cases, the hostile act begins not when someone crosses a fence or opens a door, but when a vehicle begins behaving differently.

This is where the future of perimeter security begins.

The next evolution of LPR is not better plate recognition. It is Vehicle-Centric Situational Awareness (VCSA)—the continuous collection, correlation, and analysis of vehicle identity, behavior, movement, and operational context to assess intent, predict risk, and enable proactive security decisions.

The vehicle itself is becoming an intelligent sensor.

From Identification to Intent

Traditional LPR systems operate much like access control systems. A camera reads a license plate, compares it against a watch list, and generates an alert if there is a match. The process is transactional and reactive.

Vehicle-Centric Situational Awareness shifts the objective from identifying a vehicle to understanding its intent.

Instead of asking, “Who owns this vehicle?” security teams begin asking a different set of questions.

• Why has this vehicle returned four times this week?
• Why has it approached three different gates but never entered?
• Why did it remain parked outside the employee entrance for thirty minutes after midnight?
• Why has it appeared at three distribution centers operated by the same company over the past month?
• Why is it avoiding the designated truck entrance?

These are operational questions rather than investigative ones, , reflecting a fundamental shift from identifying vehicles to understanding their behavior and intent.

Recognizing Hostile Vehicle Behaviors

Just as modern video analytics identify suspicious human activity such as loitering, intrusion, or abandoned objects, the next generation of vehicle intelligence is learning to recognize hostile vehicle behaviors.

Examples include repeated drive-by’s of executive entrances, vehicles lingering near loading docks without conducting business, repeated failed gate access attempts, unusual after-hours arrivals, convoy behavior involving multiple vehicles, tailgating through controlled access points, or vehicles that systematically visit multiple facilities within the same enterprise.

Individually, these events may appear harmless. Viewed together, they form a behavioral pattern that may indicate reconnaissance, workplace violence planning, or coordinated criminal activity.

The objective is no longer to document a vehicle’s presence. It is to understand what that presence may reveal about intent.

A New Layer in the Perimeter Security Stack

Vehicle intelligence is not a replacement for traditional perimeter technologies. Instead, it becomes the earliest layer in a broader situational awareness architecture.

Imagine a modern logistics campus.

Long before a vehicle reaches the front gate, roadside cameras identify its make, model, color, direction of travel, and previous visit history. AI analytics compare the vehicle’s behavior against normal traffic patterns. Security software evaluates the visit against scheduled deliveries, employee parking assignments, visitor registrations, and historical activity.

If the vehicle exhibits elevated risk characteristics, cameras automatically begin tracking it. Access control policies may temporarily restrict entry. Remote video monitoring operators receive a contextual alert instead of a generic notification. Autonomous patrol robots or security officers can be directed toward the area before occupants exit the vehicle.

By the time the vehicle reaches the gate, security personnel understand the vehicle’s context, behavior, and potential risk—not simply its license plate number.

Artificial Intelligence Changes the Equation

Artificial intelligence is accelerating this evolution by enabling systems to recognize vehicles even when license plates are obscured or unavailable.

Modern analytics increasingly identify vehicles based on body style, color, damage patterns, roof accessories, trailers, commercial branding, wheel configurations, and other visual characteristics. Even when criminals swap or obscure license plates, the vehicle itself often remains recognizable.

AI also enables behavioral analysis that extends beyond individual incidents. Machine learning models can identify recurring travel routes, abnormal stopping behavior, unusual visit frequencies, convoy relationships, and route deviations that would be nearly impossible for human operators to recognize consistently.

The result is a richer understanding of vehicle behavior and intent rather than simple vehicle identity.

Early Examples Already Exist

Although Vehicle-Centric Situational Awareness is a relatively new concept, many of its building blocks are already appearing across the security industry.

Retail organizations combating organized retail crime increasingly track suspect vehicles across multiple locations rather than focusing solely on individual offenders.

Distribution centers are beginning to monitor truck dwell times, unauthorized route deviations, and repeated after-hours visits to identify cargo theft risks.

Critical infrastructure operators are beginning to use vehicle analytics to help distinguish legitimate contractors from unauthorized visitors before they reach sensitive facilities.

Smart city deployments combine roadside cameras, video analytics, and transportation data to reconstruct vehicle movement following major incidents, demonstrating how behavioral intelligence can support both public safety and enterprise security.

Together, these examples illustrate the early stages of a broader shift toward behavior-based vehicle intelligence.

The Future Is Predictive

The perimeter security industry has spent decades investing in technologies designed to detect intrusions after they occur. Vehicle-Centric Situational Awareness represents a fundamental shift toward predictive security by using AI, remote video monitoring, autonomous security systems, advanced analytics, and cloud-based security operations to recognize hostile vehicle behavior before a threat reaches the perimeter.

Rather than simply determining whether a license plate matches a watchlist or authorized vehicle database, VCSA analyzes how a vehicle behaves—its movement patterns, frequency of visits, route deviations, loitering, and other contextual indicators—to assess intent and identify potential risks early.

This evolution transforms License Plate Recognition from a reactive identification tool into a source of operational intelligence. The future of LPR is no longer about reading license plates more accurately; it is about understanding vehicle behavior as one of the earliest indicators of operational risk.

For the next generation of perimeter security, the most valuable intelligence may arrive on four wheels—long before an alarm is triggered—giving security teams the time and context needed to make proactive, informed decisions.

Building the Intelligent Perimeter

How microwave, infrared, fiber optic, and LiDAR technologies work together to improve perimeter detection.

Protecting a perimeter is no longer simply a matter of detecting whether someone crossed a fence line. Today’s security leaders want to know what triggered an alert, where it occurred, whether it represents a legitimate threat, and how quickly security teams can respond.

To meet those demands, organizations are increasingly deploying a combination of perimeter detection technologies, including microwave sensors, infrared sensors, fiber optic sensing systems, and LiDAR. Each technology offers distinct strengths and limitations, making sensor selection less about choosing a single solution and more about building a layered detection strategy.

The question facing today’s security leaders is no longer which sensor technology is best. Instead, the challenge is determining which technologies work best together to create the most effective detection strategy for a specific facility, threat profile, and budget.

Layered Detection

Successful perimeter security is built around a layered model that combines physical barriers, detection sensors, video verification, remote monitoring, and automated response technologies. An integrated technology stack continuously monitors the perimeter and provides operators with actionable information when an event occurs.

Organizations are under increasing pressure to detect threats earlier, reduce false alarms, and improve response times. At the same time, logistics operators face rising cargo theft, while utilities, data centers, and transportation providers must secure large, geographically dispersed facilities with limited staffing resources.

The result is growing demand for perimeter detection technologies that can deliver earlier, more accurate threat identification.

Microwave Sensors: Proven Long-Range Detection

Microwave sensors remain one of the most widely deployed technologies for protecting large commercial and critical infrastructure perimeters.

These systems create an invisible electromagnetic detection field between transmitter and receiver units. When an object enters or disrupts that field, the system generates an alarm.

Microwave sensors are particularly well-suited for long perimeter runs where operators require reliable detection across open areas. Utility substations, solar farms, logistics yards, and industrial facilities frequently rely on microwave detection because the technology performs well in rain, fog, dust, and other environmental conditions that can challenge optical sensors.

Microwave technology’s primary advantage is reliable long-range detection at a relatively low cost. However, microwave sensors generally provide limited information about what triggered the alarm. While they can detect movement effectively, they typically cannot distinguish between a person, vehicle, or large animal without assistance from integrated cameras or analytics.

For this reason, microwave systems are most commonly paired with thermal cameras, PTZ cameras, and video management systems that provide visual verification.

Infrared: The Workhorse of Commercial Perimeter Security

Infrared remains one of the most widely used perimeter detection technologies, particularly in commercial environments.

Active infrared systems create invisible beams between transmitter and receiver units positioned along a protected boundary. When those beams are interrupted, an alarm is generated.

The simplicity of infrared technology has made it a popular choice for commercial properties, educational campuses, healthcare facilities, warehouses, and office complexes. Lower installation costs make infrared an attractive option for organizations seeking reliable intrusion detection without a significant capital investment.

Despite its popularity, infrared technology has limitations. Heavy fog, snow, rain, and environmental contamination can affect performance. Long perimeters often require multiple beam sets and periodic maintenance to ensure proper alignment and operation.

Nevertheless, infrared remains a cost-effective option for facilities with predictable environmental conditions and limited budgets.

Fiber Optic Sensing: Turning the Fence into a Sensor

Fiber optic perimeter sensing has become an increasingly popular option for organizations protecting large and complex perimeters.

Unlike traditional point sensors, fiber optic systems transform the perimeter itself into a continuous sensing element. Fiber cable can be attached directly to fences, buried underground, or installed along critical infrastructure assets. Vibrations and disturbances create changes within the optical signal that are analyzed to detect potential threats.

Modern fiber optic systems can identify fence climbing, fence cutting, digging activity, footsteps, vehicle movement, and other forms of intrusion. More importantly, they can often pinpoint the exact location of an event along miles of protected perimeter.

This capability has made fiber optic sensing particularly attractive for airports, rail corridors, utilities, military installations, data centers, oil and gas facilities, and correctional institutions.

Fiber optic systems typically require higher upfront investment and more sophisticated design than microwave or infrared technologies. However, organizations protecting large perimeters frequently find that the precision, scalability, and operational benefits justify the additional cost.

For many critical infrastructure operators, fiber optic sensing has become a cornerstone technology for perimeter protection.

LiDAR: Bringing Intelligence to the Perimeter

LiDAR is gaining traction in perimeter security because of its ability to create highly detailed, three-dimensional awareness of the environment.

LiDAR systems emit laser pulses and measure their return time to build highly accurate spatial maps. Rather than simply detecting motion or vibration, LiDAR continuously analyzes the environment and tracks objects moving through it.

The result is richer contextual information about what is moving through a protected area, where it is going, and how it is behaving.

Modern LiDAR systems can distinguish between people, vehicles, and animals while tracking movement patterns in real time. This dramatically reduces nuisance alarms and provides security operators with significantly more information when assessing potential threats.

LiDAR is increasingly being deployed at airports, ports, logistics facilities, data centers, utilities, and other high-security environments where detection accuracy is paramount.

Beyond perimeter detection, LiDAR is becoming a foundational technology for autonomous security operations. Security robots, autonomous drones, AI-driven analytics platforms, and digital twin environments frequently rely on LiDAR-generated data to navigate, classify threats, and automate response workflows.

While LiDAR remains one of the more expensive perimeter detection technologies, its ability to provide highly detailed environmental awareness is driving adoption in a growing number of high-security environments.

The Move Toward Sensor Fusion

One of the most significant developments in perimeter security is the growing adoption of sensor fusion architectures.

Rather than relying on a single detection technology, organizations increasingly deploy multiple sensors that complement one another.

A utility substation may combine microwave detection with thermal cameras. A data center may deploy fiber optic sensing along the fence line while using LiDAR to monitor vehicle approaches. An airport may integrate radar, LiDAR, fiber optics, and video analytics into a unified command platform.

This layered approach has several advantages. Detection confidence increases, false alarms decline, and operators receive richer information about potential threats. Security teams can identify an intrusion earlier, classify it more accurately, and initiate a response faster.

The result is a more effective security operation without requiring additional personnel.

Buying for Outcomes, Not Sensors

As organizations evaluate perimeter detection technologies, the focus should remain on operational outcomes rather than individual sensor specifications.

The most successful deployments begin by identifying the threats that must be addressed. Is the primary concern fence climbing, vehicle intrusion, cargo theft, copper theft, vandalism, or unauthorized access? How large is the perimeter? What environmental conditions must be considered? How quickly must security teams respond?

Equally important is understanding how those detection systems will integrate with the broader security ecosystem. The future of perimeter security is not defined by any single technology. It is defined by how multiple technologies work together to create a continuously monitored, intelligent, and responsive perimeter.

Microwave, infrared, fiber optic, and LiDAR sensors each contribute unique capabilities to that mission. As organizations modernize their security operations, the most effective solutions will be those that combine these technologies into integrated systems that detect threats earlier, reduce operator workload, and deliver actionable intelligence before an incident becomes a crisis.

How new technologies are improving checkpoint speed, accuracy and the visitor experience

Security checkpoints have traditionally forced organizations into a difficult tradeoff: stronger detection often meant slower entry. As facilities look to improve both security and visitor experience, manufacturers are rethinking how weapons detection systems identify threats while keeping people moving.

That balancing act has driven significant innovation in gun detection and weapons screening technology. While effective threat detection remains the primary objective, the latest systems are also designed to reduce bottlenecks, simplify operations and improve the overall screening experience.

Footwear Presents a Unique Screening Challenge

Footwear continues to present one of the more persistent challenges for high-security screening environments.

Many work boots, safety shoes and even everyday footwear contain steel shanks, reinforced toes or other metal components capable of triggering alarms despite posing no security risk. These nuisance alarms often require manual secondary inspections that slow checkpoint operations and divert security personnel from higher-value tasks.

To address this challenge, manufacturers have introduced specialized shoe screening technologies capable of distinguishing common footwear materials from concealed metallic weapons. By reducing unnecessary inspections, these systems can help improve checkpoint efficiency while maintaining rigorous screening standards.

New Technologies Continue to Expand Screening Options

Recent product introductions illustrate how manufacturers are addressing both security and operational efficiency.

CEIA USA recently announced the commercial availability of its OPENGATE® 2.0 walk-through weapons detection system and SAMD® Shoe Metal Detector. According to the company, OPENGATE 2.0 combines multi-caliber weapons detection with higher throughput by reducing nuisance alarms, minimizing divesting requirements and allowing people to pass through screening with backpacks, purses and footwear.

The system also incorporates integrated shoe screening, LTE and Wi-Fi connectivity, rapid deployment capabilities and support for ASTM F3566-22 security screening standards. The companion SAMD system is designed to identify weapons concealed in footwear while distinguishing benign shoe metal components that commonly trigger false alarms.

As organizations continue modernizing entrance security, the conversation has expanded beyond whether a system can detect a weapon. Increasingly, security professionals are evaluating how technology can strengthen protection while supporting faster throughput, reducing operational burdens and creating a smoother screening experience for everyone entering the facility.

For many organizations, perimeter security is approached in separate layers: the fence line, the outer property, and facility entry points. Each area is often managed with a variety of technologies, location-specific assigned personnel, and different operating procedures. The challenge is that threats don’t move in defined systematic phases.

An unauthorized individual, suspicious vehicle, or coordinated intrusion attempt moves continuously through every layer of perimeter protection. That is why effective perimeter security is not just about strengthening individual barriers, but about ensuring every security zone works together as one connected operational system. This includes protections on secured interior areas within a facility.

The Outer Perimeter Sets the Foundation

The outer perimeter is an organization’s first line of defense to deter, detect, and prevent potential threats before they reach critical infrastructure or occupied spaces. Typically, this zone incorporates a number of security technologies and devices including any combination of fence and gate-line security, bollards and barriers, video surveillance, anomaly detection analytics, detection sensors, intrusion/alarm systems, and guard stations. The sooner suspicious activity is detected, the more time security personnel have to assess whether a threat exists and respond accordingly. However, detection alone is not enough.

Perimeter protections that only identify a threat but cannot support a coordinated downstream response create only the illusion of security. If there is no visibility into where an individual or vehicle is moving or no operational process connecting detection to enforcement, organizations risk losing control once the threat advances beyond the property.

Modern perimeter protection increasingly relies on identity-based security strategies that verify users before granting access. Technologies such as biometric identity authentication, intelligent access control, and anti-tailgating systems help organizations shift from reactive security models to proactive threat prevention.

Much of this friction is caused by how traditional access control systems are deployed. SMBs have forgone installation due to the high costs of deployment and management for electronic access control. Larger multi-site businesses also find it difficult to measure return on investment (ROI) and maintain standardization across locations. In both cases, ROI concerns have led many organizations to defer access control modernization, even as risk and operational pressure continue to increase.

Physical Enforcement Starts at the Perimeter

From the corporate campus to the hospital, data center, industrial complex, and government facility, your perimeter is the first line of defense, keeping outsiders from getting in. When someone penetrates that perimeter, your people, assets, and operations are at risk.

Layer 1: The Perimeter

The first objective is to keep unauthorized pedestrians and vehicles outside the inner fence line.

Early turnstiles were susceptible to piggybacking, forcing several people through the opening with one credential. Modern outdoor-rated sensor technology can detect piggybacking and automatically lock down the turnstile.

Many systems also include walk-away detection, which locks the turnstile if an authorized individual backs out before completing entry. It also includes rotation tracking that confirms each authorized user passes fully through the turnstile, providing more accurate occupancy records and audit trails.

Layer 2: The Building Entrance

Once inside the outer perimeter, building entrances must ensure that only authorized employees, contractors, visitors, or customers gain access to the facility.

Revolving doors used for security purposes help improve throughput while still enforcing entry. Sophisticated sampling algorithms allow only one person per “swipe” to enter through the revolving door.

High-risk facilities can also use vandal-proof or bulletproof glazing to help protect against brute force attacks.

Layer 3: Interior Transition Zones

Physical enforcement should continue inside the building as individuals move between public and restricted areas.

Entrance lobbies, elevator lobbies, and transitions between interior spaces are all potential security weak points. Businesses can protect their perimeter without restricting access to the building entry.

Optical turnstiles, security revolving doors, and controlled transitions allow traffic to move freely when authorized and prevent visitors from bypassing front desks.

Layer 4: Critical Infrastructure

Highly sensitive spaces such as server rooms, research labs, executive suites, and critical infrastructure require the strictest physical security measures.

Interlocking mantrap portals enforce one-person authentication through staged credential validation and biometric verification.

This process helps eliminate user substitution, collusion and unauthorized piggybacking.

Why Modern Threats Are Exposing the Enforcement Gap

Enterprises are experiencing growing security challenges from open floor plans, increased density, employee turnover, hybrid operations, coordinated attacks and increased contractor activity.

Meanwhile, ease of use has silently eroded security practices in many organizations.

Open doors, shared keys/padlocks, unlocked side entrances, free flow through lobbies, and reliance on cameras have become the norm.

Modern threats are exposing a hard truth: monitoring alone is not enough. Facilities need infrastructure that actively enforces security decisions before unauthorized movement occurs.

This is particularly important in regulated industries where perimeter failures can create serious consequences related to compliance, safety, operational continuity, and liability.

Security and Throughput Must Work Together

Facilities must balance secure entry with efficient movement for employees, visitors, and deliveries. The challenge is balancing security, safety, speed, and the user experience.

The most effective facilities design entrances to support flow instead of fighting against it.

Today’s converged security systems also enable organizations to link physical enforcement with operational intelligence. They can cross-reference access control events with occupancy counts, visitor management, AI-driven analytics, and video surveillance to help balance security with throughput.

Traffic patterns can be analyzed to identify unusual activity, pinpoint congestion, and even suggest corrective routing recommendations, all in real time, with AI-driven solutions.

Biometrics are another way technology is enhancing the security/user experience balance. Facial recognition, iris scans, and fingerprint authentication can help organizations increase identity assurance while also enabling faster, touchless experiences.

The Future of Layered Security is Coordinated Physical Enforcement

The future of layered security won’t be only about adding more technology, but also about focusing on collaborative ecosystems that tie detection to decisions, monitoring, and physical enforcement.

Entrances, barriers, portals, turnstiles, and other controlled transition points are both enforcement and active elements of your total security solution that put policy into physical action.

Pre-deployment testing is becoming increasingly important as integrated security ecosystems grow more complex.

The goal is no longer isolated security technologies operating independently, but coordinated physical enforcement throughout the entire facility journey.

Why Physical Enforcement Is Essential to Layered Security

Credentials, analytics, surveillance systems, and cybersecurity platforms are all essential components of modern layered security. But they are only part of the equation.

Access control and analytics can identify threats, but only physical enforcement can stop unauthorized movement.

The effectiveness of every layer of security ultimately depends on whether access decisions can be enforced physically, from the perimeter fence to the most sensitive interior spaces.

Access control and physical infrastructure will converge in the security design of future-proof organizations. Entryways, barriers, sensors, analytics, and identity authentication will operate as a unified ecosystem that safeguards people, assets, and operations throughout the facility journey.

By Kurt Measom, Boon Edam

Building reliable connectivity where traditional networks end

Perimeter security technology has advanced dramatically over the past decade. AI-powered video analytics, autonomous security robots, remote video monitoring (RVM), intelligent fence sensors, LiDAR, radar, thermal cameras, access control, and cloud-based command centers have transformed how organizations protect critical assets.

But none of these technologies can deliver their full value without reliable communications.

For many perimeter applications, communications infrastructure has become the single most important design consideration. Whether protecting a utility substation, construction project, rail yard, pipeline, solar farm, mining operation, border facility, or temporary event, security systems increasingly operate in locations where fiber and wired broadband do not exist.

Today’s security integrator is often asked a question that was once outside the scope of physical security:

“How are we going to connect to this site?”

The answer has evolved from “We can’t connect that site,” to “We can connect it, but it’s too expensive,” to “Today, multiple communications options make reliable connectivity practical and affordable. “What was once considered impossible—and later prohibitively expensive—is now achievable through multiple reliable and increasingly affordable communications technologies.

Communications Is Now Part of the Security Design

Many of today’s perimeter deployments are remote, temporary, mobile, geographically distributed, and located outside enterprise network boundaries.

Examples include construction sites, utility substations, renewable energy facilities, rail infrastructure, remote communications towers, oil and gas facilities, equipment yards, ports, agricultural operations, wildfire detection systems, and mobile security trailers.

For these applications, communications often determine whether advanced security technology can be deployed at all.

The communications architecture must support not only video transmission, but also AI analytics, health monitoring, firmware updates, access control events, audio communications, environmental sensors, and sometimes autonomous robotic systems.

Bandwidth, latency, resiliency, and operating cost have become just as important as camera resolution.

Cellular Has Become the Default Choice

For most remote perimeter projects, cellular connectivity has become the default communications option.

The rollout of 4G LTE and increasingly 5G has dramatically expanded coverage while lowering deployment costs.

Today’s industrial cellular routers support multiple carriers, automatic carrier failover, private APNs, VPN encryption, edge computing, remote device management, and dual-SIM redundancy.

Cellular deployments offer many advantages. They are fast to install, require no trenching or fiber infrastructure, offer predictable monthly operating costs, and generally provide sufficient bandwidth for AI-assisted video applications using modern compression technologies.

For construction sites, mobile surveillance trailers, temporary events, retail parking lots, and remote substations, cellular has become the communications backbone.

Its primary limitation remains coverage, as many critical infrastructure locations simply fall outside reliable cellular service. Even where coverage exists, network congestion during disasters or major public events can reduce performance precisely when security becomes most important.

Satellite Communications Has Entered a New Era

Until recently, satellite communications were viewed as the communications option of last resort for remote perimeter deployments.

Low Earth Orbit (LEO) satellite networks have fundamentally altered expectations.

Unlike traditional geostationary satellites positioned approximately 22,000 miles above Earth, LEO constellations operate hundreds of miles overhead, dramatically reducing latency while increasing throughput.

For remote security applications, this makes continuous monitoring practical in locations that were previously difficult or impossible to connect.

Sites previously considered impossible to monitor continuously—including isolated utility assets, border infrastructure, mining operations, renewable energy farms, forestry operations, pipelines, and offshore facilities—can now support high-quality video, cloud-based management, AI-assisted monitoring, and remote diagnostics.

The primary tradeoff remains cost. While pricing continues to decline, satellite services still carry higher recurring expenses than cellular and require an unobstructed view of the sky. For many remote sites, however, satellite is evolving from a technology of last resort into the preferred long-term communications solution.

Private Wireless Networks Are Growing

Some organizations prefer to own their communications infrastructure.

Private wireless networks—including licensed microwave links, point-to-point radio systems, mesh networking, and private LTE or 5G deployments—are becoming more common across large industrial campuses.

Utilities, ports, airports, military installations, mining companies, and large manufacturers often deploy their own wireless infrastructure to maintain complete operational control.

Private wireless networks offer predictable performance, eliminate recurring carrier fees, provide greater cybersecurity control through network segmentation, deliver very low latency, and offer high reliability for mission-critical applications.

The tradeoff is higher upfront capital investment and greater engineering complexity.  These systems are best suited for organizations connecting dozens or hundreds of remote assets over the long term.

Radio Frequency Still Plays an Important Role

While broadband communications receive most of the attention, traditional radio frequency (RF) technologies remain essential components of perimeter security.

Long-range RF systems continue to support intrusion sensors, emergency communications, gate controls, environmental monitoring, backup alarm reporting, and remote command functions.

Because many RF systems require minimal bandwidth, they remain reliable under adverse weather conditions and can operate for years using battery or solar power. Rather than competing with broadband communications, RF increasingly serves as a resilient backup communications path.

Mesh Networking Extends Coverage

Wireless mesh networking has become an effective solution for large campuses where extending fiber would be prohibitively expensive.

Rather than connecting every device directly to the internet, each node relays communications through neighboring devices until reaching a network gateway.

Mesh architectures are particularly well suited for solar farms, utility campuses, industrial facilities, smart fencing, environmental sensor deployments, and large logistics yards.

If one communication path fails, data automatically reroutes through another node, improving overall network resilience.

The Rise of Hybrid Communications

Increasingly, security designers combine multiple communications technologies rather than relying on a single network.

A modern perimeter deployment may use fiber as the primary connection, cellular backup, satellite failover, RF alarm reporting, and mesh networking to connect distributed sensors.

Software-defined networking can automatically select the most efficient communication path based on network conditions. This multi-layer approach significantly improves resilience during storms, carrier outages, construction damage, or cyber incidents. Communications redundancy is becoming as important as power redundancy.

Communications Must Support AI

One of the biggest changes in remote security is the sheer volume of data now generated at the perimeter.

High-resolution cameras, AI analytics, video verification, autonomous patrol robots, drones, license plate recognition, and sensor fusion all generate significantly more information than legacy intrusion systems.

Sending every video stream continuously to the cloud is rarely practical. Instead, edge computing has become a critical part of communications design. Processing AI locally reduces bandwidth requirements by transmitting only events, metadata, alarms, or selected video clips. This approach lowers communications costs while improving response times.

For remote sites operating over cellular or satellite, edge AI is rapidly becoming essential rather than optional.

Questions to Ask Before Designing a Communications Architecture

Selecting the right communications architecture requires in-depth evaluation and information gathering. Here are some questions to help you navigate the process:

• How much bandwidth will our cameras, analytics, and connected devices require? What happens if the primary communications path fails?
• Will the system continue operating during disasters?
• Is latency important for real-time intervention?
• Can AI be processed locally to reduce bandwidth?
• Are recurring communications costs sustainable over five to ten years?
• How will remote firmware updates and cybersecurity patches be delivered?
• Will the communications platform support future technologies such as robotics, drones, and additional sensors?
• How is network traffic encrypted and monitored?
• How does our cyber policy impact our options?
• Can the communications infrastructure be scaled as new sites are added?

Begin with the End in Mind

Perimeter security begins with connectivity, not endpoints.

The rapid expansion of cellular coverage, the emergence of LEO satellite services, advances in private wireless networking, edge AI, and hybrid communications architectures have given organizations more choices than ever before.

As perimeter systems become more autonomous, more intelligent, and more distributed, communications is evolving from a supporting technology into one of the foundational layers of modern security architecture.

Ultimately, even the most sophisticated perimeter security system is only as effective as the communications network that supports it.