A Technology Selection Guide

Beyond the Spec Sheet
The Critical Infrastructure Guide to Perimeter Detection
This guide is for senior security buyers, consultants, engineers, systems integrators, specifiers, and procurement teams responsible for outdoor perimeter protection and PIDS evaluation at substations, data centers, airports, ports, water facilities, oil and gas sites, and other high-consequence environments.

Download PDF Version

Published by SightLogix | Princeton, NJ | www.sightlogix.com
Introduction
Most perimeter security failures are purchased long before they happen.
They begin when teams compare products by range, resolution, frame rate, or AI claims instead of asking the field questions that matter: Will the system detect a real intruder early enough to respond? Will it ignore wildlife, wind, rain, headlights, shadows, vegetation, and nuisance motion? Will operators still trust the alerts after 90 days? [S1] [S11]
This guide is about evaluating perimeter security beyond the spec sheet.
Thermal cameras, visible cameras with analytics, fence sensors, buried cable, radar, LiDAR, and drones all have a role. They do not solve the same problem. Some monitor close-in areas. Some confirm boundary contact. Some provide long-range awareness. Some support response. The hard part is deciding which technology should carry primary detection for your site.
One requirement runs underneath all of them: An alert only matters if an operator can see what triggered it and respond correctly. A sensor that reports contact but can't show whether it was a person, an animal, or the wind leaves the operator guessing at the worst moment. The strongest architectures pair detection with video to see the threat — and the detection layer itself is most valuable when it does both.
SightLogix publishes this guide from a clear point of view: for many critical infrastructure sites, the hardest problem is reliable mid-range outdoor detection with low nuisance alarms. But no single technology is right for every zone or site. The goal is to show where each category fits, where it struggles, and what buyers should prove before procurement. [S17]
For more than 20 years, SightLogix has been solving the hardest problems in outdoor perimeter detection — early warning at range, precise alert location, and reliable low-nuisance performance — for high-consequence sites where operational risk is highest.

PIDS Technology Selection Guide | © 2026 sightlogix.com

2

Executive Summary
Perimeter security decisions become clearer when the site is evaluated by detection zone, response requirement, and false alarm tolerance rather than by product category or headline specification.
Mid-Range Zone is Critical
For many high-consequence sites, the critical zone is the mid-range outdoor perimeter: roughly 50 to 500 meters where early warning is possible and necessary. [S6] [S7]
Visible Cameras Have Limits
Visible cameras remain valuable for access points, close-in coverage, verification, and documentation, but are limited as the primary detection layer for large outdoor perimeters because they depend on light and favorable visibility. [S15]
Fence Sensors Confirm, Don't Warn Early
Fence-mounted and buried sensors provide useful boundary confirmation, but most detect at or near the fence line, after response time has narrowed. [S1] [S6] [S7]
Radar and Drones Complement
Radar and drones can be powerful complements for large or complex sites, but usually need a camera-based layer for classification and confirmation. [S3] [S4] [S5] [S9]
Thermal Leads Mid-Range
For 50-500 meter outdoor perimeters, purpose-built thermal detection is usually the first category to evaluate because it detects emitted heat rather than reflected light and can perform in darkness, glare, and difficult weather when proven on site. [S12] [S17]

PIDS Technology Selection Guide | © 2026 sightlogix.com

3

A Technology Selection Guide

1
Four Metrics That Actually Matter

Four Metrics That Actually Matter
Choosing a perimeter security technology is easier when the evaluation criteria are set before the comparison begins. Most assessments start with specifications: sensor resolution, frame rate, field of view, detection range, and AI claims. Those numbers may be useful, but they rarely tell you whether the system will perform on your perimeter.
These four criteria apply to every category in this guide — thermal, visible, radar, fence and buried sensors, LiDAR, and drones. What changes is how each technology answers them.
1
1
Detection Range at Reliable Probability of Detection
The distance at which the system generates a reliable, classified alert in real site conditions — not theoretical maximum range.
2
2
False Alarm Rate in Real Outdoor Environments
Performance in wind, vegetation, animals, rain, shadows, headlights, and seasonal variation — not in a controlled demonstration.
3
3
Reliability Across Site Conditions
Consistent performance in darkness, glare, fog, rain, snow, wind, thermal contrast, and vibration — validated against your specific site.
4
4
Total Cost of Operation
Five-year cost per meter protected, including poles, cabling, trenching, networking, false alarm response, maintenance, and hardware replacement.

PIDS Technology Selection Guide | © 2026 sightlogix.com

5

The Spec-Sheet Trap
Most vendor conversations start with the claims in the left column. Understanding what those claims actually mean and what to ask instead is where a useful evaluation begins.

PIDS Technology Selection Guide | © 2026 sightlogix.com

6

Metric 1: Detection Range at Reliable Probability of Detection
Not the theoretical maximum range at which a sensor can see a target. The relevant number is the distance at which the system generates a reliable, classified alert under the worst conditions your site actually presents.
Detection distance isn't always measured the same way. Every sensor has an easy case it's measured against and a harder one it will actually face, and published range sometimes reflects the easy one. This can differ by technology: radar, for example, is strongest when someone is approaching the sensor, weaker when someone moves across its field of detection. The example below shows the video case: a person crossing the field of view is easy; someone walking straight toward the camera is hard — less motion, a smaller changing target. A system specified for that harder inbound case states a range you can actually count on, even if it looks smaller than one measured the easy way.


PIDS Technology Selection Guide | © 2026 sightlogix.com

7

Metric 2: False Alarm Rate in Real Outdoor Environments
False alarms are not a minor operational inconvenience. Each requires a response. Over time, a system that produces regular false alarms trains operators to discount alerts — which is precisely the failure mode a perimeter security system is supposed to prevent. [S11]
The relevant measure is not false alarm rate in a controlled demonstration. It is false alarm rate under real site conditions: wind-driven vegetation, animals, rain, atmospheric change, vibration, shadows, and moving debris across hundreds of meters of open perimeter. [S1] [S15]
The foundational challenge of outdoor automated detection is simple: perimeter systems are designed to detect movement, and outdoors, everything moves.
A manufacturer should be able to explain in plain terms how their system distinguishes a real intruder from environmental noise — not with vague references to "better algorithms," but with a specific account of the detection architecture and why it produces low false alarm rates across real deployments.

PIDS Technology Selection Guide | © 2026 sightlogix.com

8

Metric 3: Reliability Across Site Conditions
A system that performs well in favorable conditions and degrades in difficult ones provides less protection precisely when protection is most needed. Thermal detection is generally more resilient than visible-light cameras in darkness, glare, and many low-visibility conditions - but performance should still be validated against the specific rain, fog, snow, wind, thermal contrast, and vibration conditions the site presents. [S13] [S14] [S15]
Within any technology category, real-world performance in cold, wind, rain, and vibration varies considerably depending on how the system was designed and tested. [S13] [S14] [S15]

PIDS Technology Selection Guide | © 2026 sightlogix.com

9

Metric 4: Total Cost of Operation
Unit price is one of the least useful numbers in a perimeter security evaluation. The relevant figure is five-year cost per meter protected — poles, cabling, trenching, networking, false alarm response, and maintenance included. [S42]
A system that detects just as reliably at longer range simply needs fewer devices to cover the same perimeter. The infrastructure savings alone can reduce project costs by a third or more.

PIDS Technology Selection Guide | © 2026 sightlogix.com

10

A Technology Selection Guide

2
Knowing Where You Need Coverage

Three Detection Zones – Knowing Where You Need Coverage
Before comparing technologies, define which zones of your perimeter you need to protect. A system optimized for 30 meters offers little value at 200 meters — and vice versa. Knowing your zones narrows the field significantly. Large sites require coverage across multiple zones. [S6] [S7] [S42]
Short Range
0 to 50 Meters
Visible cameras, close-in sensors, and LiDAR. Best for access points, building perimeters, gate approaches, and tight spaces where monitoring matters more than early warning.
Mid Range
50 to 500 Meters
The most consequential zone for critical infrastructure: open ground between the boundary and critical assets. Detection distance is response time — a mid-range alert gives operators time to act. For data centers, substations, and similar sites, this is the primary security challenge.
Long Range
1,000 Meters and Beyond
Radar and specialized long-range systems. Airports, large utility campuses, and ports requiring early warning well before the boundary. Typically complements mid-range detection rather than replacing it.

PIDS Technology Selection Guide | © 2026 sightlogix.com

12

A Technology Selection Guide

3
Technology Categories — Where Each Fits

Technology Categories — Where Each Fits
The following comparison applies the four evaluation metrics from Section 1 to each technology category. Each is assessed on its own terms - what it does well, where it belongs, and what its honest limitations are for large outdoor perimeter applications at critical infrastructure sites. Some categories compete directly for the same application. Others serve different zones and work best alongside other technologies in a layered deployment.
The manufacturers referenced throughout this section are examples of companies commonly evaluated within each technology category. These lists are not exhaustive, and inclusion does not imply that every product from each manufacturer is appropriate for every application. The right solution for any site depends on perimeter geometry, required detection range, lighting conditions, weather, terrain, integration requirements, regulatory constraints, budget, and the operational consequences of a missed detection or nuisance alarm. [S42]

PIDS Technology Selection Guide | © 2026 sightlogix.com

14

Thermal Security Cameras with Analytics
Thermal cameras detect emitted heat rather than reflected light, which gives them a major advantage where visible cameras struggle: darkness, glare, lighting transitions, and many low-visibility outdoor conditions.
Because the image that raises the alert is also the image the operator sees, a thermal camera provides detection and visual confirmation in one device — in the darkness, glare, and weather where a visible camera would show nothing.
For mid-range perimeter protection, roughly 50 to 500 meters, thermal detection is often the right starting point. This is especially true where early warning, low nuisance alarms, and reliable performance across changing outdoor conditions are critical.
Manufacturers to evaluate in this category:
Purpose-built
SightLogix is purpose-built for thermal perimeter intrusion detection, with edge-based analytics, geospatial target tracking, onboard stabilization, and long-range detection — engineered for reliable performance at critical perimeters. [S17]
Broad-range
FLIR and Axis are broad, established thermal camera manufacturers with large product portfolios and wide compatibility across video management platforms. Both offer outdoor thermal cameras appropriate for a range of applications, where SightLogix is purpose-built for perimeter intrusion detection as the primary design objective.


PIDS Technology Selection Guide | © 2026 sightlogix.com

15


Thermal Cameras: Questions to Ask
The architecture inside the camera matters: These choices determine whether a system works in the field or only looks strong in a demonstration. [S12] [S17]
How video is processed
Thermal sensors generate dense, data-intensive streams. Because most edge processors can't handle raw video, many systems compress it before analysis — discarding fine scene detail that accurate long-range detection requires.
Ask: Does the system analyze raw thermal data before compression, or does it analyze compressed video?
Geospatial Scene Registration
A camera that maps every pixel maps to the real world calculates the actual size of any object — not just how many pixels it occupies. A dog near the camera "looks" bigger than a person 200 meters away, but a geospatially registered system knows the actual dimensions to ignore the dog and alarm on the person.
Ask: Does the system geo-register the scene to calculate real-world object dimensions — or does it work from apparent pixel size alone?
Does AI Detect or Classify
AI is most effective as a classifier, not a primary detector. The better architecture uses geospatial analysis to identify candidates first, then brings AI in to classify them. When AI scans every pixel of a wide-area scene, it reduces effective range or increases missed detections.
Ask: Does the system use AI to detect, to classify, or both — and at what stage?

PIDS Technology Selection Guide | © 2026 sightlogix.com

16

Thermal Cameras: Questions to Ask (cont.)
What the AI Was Trained On
AI trained on real-world outdoor thermal imagery behaves differently in the field than AI trained on synthetic or lab data. This distinction rarely appears in marketing materials but is one of the strongest predictors of false alarm performance.
Ask: Was the AI trained on real-world outdoor imagery, synthetic data, or a combination — and under what conditions?
Dual-Spectrum Integration
Some systems combine thermal and visible cameras, running independent analytics on both. Thermal carries detection in darkness; visible adds color, context, and behavioral cues as light increases — improving discrimination where thermal contrast alone is marginal.
Ask: If both thermal and visible sensors are present, do they work together analytically in real time, or does one simply record while the other detects?
Geospatial Output and PTZ Integration
A detection event that includes precise GPS coordinates can automatically command a PTZ camera to reposition, zoom, and track the target autonomously — even beyond the detecting camera's coverage area.
Ask: Does a detection event automatically direct a PTZ to the target's location?
Onboard Stabilization
Cameras mounted on poles move — from wind, vibration, or thermal expansion. Without stabilization, the analytics can't distinguish a moving target from a moving background, and detections fail. In a head-to-head test with cameras shaking on the same pole, one system with stabilization detected at 90 meters. The other was limited to 21 meters.
Ask: Does the system include onboard stabilization — and what happens to detection performance when the camera is moving?

PIDS Technology Selection Guide | © 2026 sightlogix.com

17

Visible Cameras and Video Analytics
Visible cameras are the most widely deployed surveillance technology at critical infrastructure sites and remain appropriate for a range of applications: access control monitoring, close-in perimeter coverage, and verification and documentation of detected events. For these applications they are well-understood, widely supported by video management systems, and cost-effective.
Their limitation as the primary mid-range outdoor detection layer is fundamental: visible cameras depend on reflected light. Performance can degrade at night, in glare, in fog, in rain, during lighting transitions, and wherever IR illumination is insufficient or impractical. The right use of visible cameras in a perimeter security architecture is typically as a verification and documentation layer alongside a detection technology that performs reliably at range and in all conditions.
These companies offer established visible-camera platforms with mature onboard and platform-based analytics - including intrusion detection, object classification, line crossing, and loitering detection - and broad compatibility with enterprise video management systems.
Buyers evaluating visible cameras for outdoor perimeter use should pay particular attention to how each system performs at night, in adverse weather, at the required detection distance, and with whatever IR illumination infrastructure their site allows. [S15] The questions below provide a consistent framework for that evaluation regardless of manufacturer.

PIDS Technology Selection Guide | © 2026 sightlogix.com

18

Viable Cameras: Questions to Ask
Documented false alarm rate
The relevant figure is false alarm rate in real outdoor conditions - wind, shadows, rain, vegetation, animals - not in a controlled demonstration. A vendor confident in their outdoor performance will share it.
Ask: What is the documented false alarm rate in outdoor conditions at the distances and lighting conditions your perimeter presents?
Analytics architecture
Ask how the system distinguishes a person from a moving shadow, a rain-streaked lens, or blowing vegetation. Concrete answers - geospatial filtering, object size calculation, scene calibration - indicate a system designed for outdoor use. Vague references to "deep learning" alone do not.
Ask: How does the analytics engine handle environmental motion, and what specific mechanisms reduce outdoor false alarms?
Mid-range detection layer
For perimeters larger than roughly 50 meters, visible cameras are most effective as a verification and documentation layer alongside a technology that provides reliable detection at range in all conditions.
Ask: How does this system integrate with mid-range detection technology, and can PTZ cameras be automatically directed to a confirmed alert location?

PIDS Technology Selection Guide | © 2026 sightlogix.com

19

Fence-Based and Buried Cable Sensors
Mechanical fence sensors, fiber distributed acoustic sensing, and buried fiber optic cable systems all detect at or below the fence line. They are fence-line confirmation technologies, not early warning systems — and within that role, they are effective and appropriate for a range of sites.
Technologies range from microphonic cable and electrostatic sensing to fiber-optic distributed acoustic systems designed for long linear perimeters and high-security applications
Buried and Covert Sensors
Highly effective for fence integrity confirmation and long linear perimeters — pipeline corridors, utility rights-of-way — where a continuous sensing element along the boundary adds a reliable confirmation layer. Buried cable systems are covert and passive, with no visible above-ground infrastructure.
The Reactive Detection Reality
Many technologies in this category detect at or near the moment of boundary contact. Response time begins counting from that point — not from when an intruder started crossing open ground. For high-consequence sites, this is the core limitation.

PIDS Technology Selection Guide | © 2026 sightlogix.com

20

Fence Sensors: Questions to Ask

Localization Precision
Detection systems in this category vary significantly in how precisely they identify where along the perimeter a contact event occurred. Some provide zone-level indication; others provide GPS-level coordinates. The precision determines how quickly a response team can act.
Ask: What level of localization does the system provide — fence zone, segment, or GPS-level coordinates?
Environmental Noise Rejection
Fence-mounted sensors are susceptible to false alerts from wind, animals, debris contact, and vegetation growth against the fence line. Buried cable systems are more immune to surface environmental noise but require careful installation to avoid false events from ground vibration.
Ask: What environmental conditions produce false alerts, and what is the documented false alarm rate in outdoor deployments?
Integration with an Early-Warning Layer
The most effective deployments pair fence sensors with a technology that detects threats in the open ground before they reach the boundary.
Ask: How does this system integrate with a perimeter detection layer that provides early warning before the fence line is reached?

PIDS Technology Selection Guide | © 2026 sightlogix.com

21

Ground-Based Radar
Ground-based radar detects and locates moving objects across large areas using active RF signals. It can operate in darkness and many low-visibility conditions, and it can cover ranges well beyond what camera-based systems cover cost-effectively. For large open-area sites and long-range applications, it is a proven technology with substantial installed base at critical infrastructure facilities across North America.
Experienced integrators describe the relationship between radar and camera-based detection in straightforward operational terms: Thermal cameras protect the fence line and provide classification at the point of approach, while radar continues to track a subject once they have crossed the boundary and are moving through the interior.
Classification Requires a Camera Layer
Radar detects and tracks motion; classification claims should be tested in the paired assessment workflow. A radar alert may tell an operator that motion occurred at a set of coordinates before the operator has visual context about whether it is a person, an animal, a vehicle, or wind-driven debris.
Natural Complement to Mid-Range Thermal AI
At complex sites — airports, large open utility campuses — radar and mid-range thermal detection are routinely deployed together in layers: radar covering the long-range zone beyond 500 meters, thermal AI providing classification and verified alerts in the mid-range zone.

PIDS Technology Selection Guide | © 2026 sightlogix.com

22

Ground-Based Radar: Questions to Ask
Visual Classification Layer
Radar tells you that something is moving and where; classification claims should still be proven through the paired assessment workflow. The quality of the integration between radar and its paired visual confirmation system matters as much as the radar's detection performance.
Ask: What camera or classification system is paired with the radar, and how is alert-to-confirmation handoff managed operationally?
Coverage Zone and Cost Per Meter
Radar is most cost-effective for large open-area sites and long-range applications. For sites where the primary detection zone is under 500 meters, the cost and infrastructure of a radar deployment may exceed what mid-range thermal technology requires for equivalent or better performance.
Ask: At the detection distances your perimeter requires, how does radar cost per meter compare to mid-range thermal AI with integrated classification?
Integration with Layered Detection
At complex sites with multiple detection zones, the quality of integration between radar and thermal AI — how alerts are correlated, how confirmation workflows are structured, how operators interact with a combined picture — determines whether two systems produce a coherent security response or two independent alert streams.
Ask: How do radar alerts integrate operationally with mid-range detection and classification systems, and what does the combined operator interface look like?

PIDS Technology Selection Guide | © 2026 sightlogix.com

23

LiDAR
LiDAR provides accurate real-time 3D spatial data, works in all lighting conditions, and raises no visual privacy concerns since it captures geometry rather than imagery. For close-in applications it is a legitimate and increasingly capable technology.
Representative manufacturers to evaluate in this category include Quanergy, which offers the MQ-8 sensor with QORTEX perception software, and Blickfeld, whose solid-state QbProtect sensor is actively marketed for CIP perimeter applications and eliminates the mechanical reliability concerns of earlier spinning-sensor designs.
LiDAR remains a more specialized and actively evolving category than thermal, radar, or fence-based sensing. The manufacturer landscape reflects that — fewer established perimeter-specific players, more variation in how systems are positioned and deployed. The questions below are designed to help buyers evaluate whether LiDAR's genuine strengths apply to their specific application before committing to a deployment architecture built around it.

PIDS Technology Selection Guide | © 2026 sightlogix.com

24

LiDAR: Questions to Ask
The Range Ceiling Is Physics
Reliable outdoor intrusion detection with LiDAR is typically strongest at short range, often around 50 to 100 meters depending on site conditions. Rain, fog, snow, atmospheric effects, and cost per meter should be tested carefully before LiDAR is considered for open mid-range perimeter coverage.
Ask: At the detection distances your full perimeter requires, how many units are needed and what is the total cost per meter of coverage?
Right Technology, Right Zone
For access control points, building entrances, and close-in zones under 50 meters, LiDAR is a strong candidate. For large open perimeters in the mid-range zone, the range ceiling and cost per meter may make it a poor primary detection tool unless the site test proves otherwise.
Ask: Is this application a close-in zone where LiDAR's strengths apply, or an open mid-range perimeter where they don't?
The Sensor is Only Part of the System
LiDAR hardware has advanced faster than the software layers that turn point-cloud data into reliable, actionable alerts. Classification accuracy, false alarm rates, and integration with existing security management platforms vary significantly between vendors.
Ask: What does the analytics and alert management layer look like, and how does it integrate with your existing VMS or security operations platform.

PIDS Technology Selection Guide | © 2026 sightlogix.com

25

Drone Patrol Systems
Drones should be evaluated in three separate roles, because each solves a different perimeter security problem.
Primary Detection Layer
Autonomous drone-in-a-box systems patrol defined areas on schedule or in response to an alarm.
Mobile Response Layer
Drones dispatched to alarm coordinates from fixed perimeter sensors, cameras, radar, or lidar.
Drone Threat Layer
Counter-drone detection and response for sites where unauthorized drones create security risk.
Autonomous drone systems are becoming a practical extension of perimeter security for commercial, industrial, government, and critical infrastructure sites. The strongest fit is not as a replacement for fixed detection, but as a mobile verification and response layer that can be dispatched to alarm locations.
Sunflower Labs is the most directly aligned with autonomous drone security, combining AI-based detection, thermal imaging, and mobile response. Skydio brings a U.S.-based autonomous platform with docked-drone capability across security, inspection, and infrastructure use cases. Percepto is more established in drone-in-a-box deployments for utilities, energy, and large industrial monitoring. Easy Aerial is especially relevant for defense, government, and high-security environments, with U.S.-manufactured, NDAA-compliant tethered and docked systems.ons.

PIDS Technology Selection Guide | © 2026 sightlogix.com

26

Drone Patrol Systems: Questions to Ask
Fixed Detection Coverage During Downtime
Weather-grounded and recharging periods can last hours. Evaluate the drone system and the fixed detection layer together, not separately.
Ask: What fixed detection system provides perimeter coverage during weather-grounded and recharging periods, and what coverage gaps exist?
GPS-Cued Dispatch Capability
The most effective integration model: a fixed system detects a target and dispatches a drone to its GPS location. The depends on how accurately the fixed system can locate a target geospatially.
Ask: Does the fixed system provide GPS coordinates in alert metadata, and how does the drone act on that data?
Regulatory and Operational Constrain
FAA regulations, airspace restrictions, and local requirements may constrain drone operations — particularly near airports, power generation facilities, and government installations.
Ask: What airspace restrictions apply to this site, and do they constrain autonomous drone patrol?

PIDS Technology Selection Guide | © 2026 sightlogix.com

27

Technology Comparison: At a Glance

PIDS Technology Selection Guide | © 2026 sightlogix.com

28

A Technology Selection Guide

4
What High-Consequence Sites Actually Require

What High-Consequence Sites Actually Require
For many sites, the right answer is a layered system — different technologies deployed at different distances and for different purposes.
At high-consequence sites — where a missed detection can affect power, water, transportation, or public safety — the primary detection layer must meet a higher standard. Three of the four evaluation metrics stop being selection criteria and become operational requirements.
Detection Must Provide Response Time
Detection must occur far enough from the boundary to give a security team time to respond — minutes, not seconds
False Alarm Rates Must Build Operator Trust
False alarm rates must be low enough that operators trust every alert — a team conditioned by false alarms is the vulnerability at the moment a real intrusion occurs.
All-Conditions Performance Is the Baseline
All-conditions performance is the baseline, not a premium feature — a patient adversary will wait for the conditions that defeat a system that only works in favorable weather.

PIDS Technology Selection Guide | © 2026 sightlogix.com

30

Infrastructure Efficiency and Long-Term Reliability
Security investments at critical infrastructure sites are capital expenditures justified over years. The relevant financial question is total cost over the life of the system: how many devices are required, what infrastructure do they demand, and what does ongoing maintenance cost?
A system that covers more perimeter per installation point reduces poles, cable runs, trench segments, and maintenance points for the life of the deployment.
Hardware reliability is not a spec to take on faith — ask for a published failure rate, not an estimated one.
Infrastructure Savings Add Up
  • Fewer poles and foundations
  • Fewer cable runs and trench segments
  • Fewer power connections
  • Fewer maintenance points over the system life

PIDS Technology Selection Guide | © 2026 sightlogix.com

31

Knowing Exactly Where, Not Just That Something Was Detected
The operational difference between "an alert triggered" and "a person is at these coordinates, here is what they look like" is significant under any threat scenario, and decisive under a coordinated one.
An alert that doesn't tell an operator precisely where an event occurred or what triggered it slows response before it starts. At a 1,200-meter substation perimeter or a large port facility, that delay matters.
Zone-Level Alert
Operator must interpret the alert, locate the event, and manually confirm what is happening. Response is delayed before it starts.
GPS-Coordinate Alert with PTZ Handoff
Geospatial coordinates in the alert metadata with automatic PTZ cueing give operators immediate situational awareness — not a starting point for investigation.

PIDS Technology Selection Guide | © 2026 sightlogix.com

32

A Technology Selection Guide

5
Real World Use Cases


What Field Deployments Show
The evaluation criteria in this guide are not theoretical. The questions around detection accuracy, false alarm performance, infrastructure efficiency, and operator trust have been tested in real deployments across the site types this guide addresses.
Independent Government Validation at an Airport Perimeter
The U.S. Transportation Security Administration commissioned an Operational Test and Evaluation through the National Safe Skies Alliance at a North American airport. Evaluators conducted more than 900 intrusion scenarios across multiple detection zones covering the facility's outer perimeter — an environment characterized by mixed topography, inconsistent illumination, and proximity to public roadways. The evaluation was conducted without vendor involvement during testing. Target tracking capability was confirmed fully functional throughout the evaluation period.
Electrical Substations: Detection Reliability and Infrastructure Efficiency
A major North American utility operating Tier 1 and Tier 2 electrical substations evaluated multiple perimeter security technologies before selecting a thermal AI detection system. The selection criteria centered on detection reliability and low nuisance alarm rate. At individual substations, between eight and twenty-eight sensors are deployed per perimeter, each paired with an auto-steered PTZ camera that responds to GPS-located alerts. The system supports NERC CIP 14 regulatory compliance and operates continuously, day and night, regardless of lighting or weather conditions.
Data Centers: Competitive Evaluation Under Real-World Conditions
A global data center operator conducted a structured competitive evaluation of multiple video analytic technologies. Each vendor was allowed to configure their system initially, then required to perform without further involvement — ensuring results reflected actual field performance rather than assisted demonstration. The thermal AI system was selected based on demonstrated detection performance, early warning capability, edge-based processing architecture, geospatial detection zones, and automatic image stabilization for outdoor environmental conditions.

PIDS Technology Selection Guide | © 2026 sightlogix.com

34

Conclusion
The framework in this guide points many critical infrastructure sites toward the same conclusion: the primary detection layer should be selected around the zone where failure matters most.
For many sites, that is the mid-range outdoor perimeter, where early warning, low nuisance alarms, all-conditions performance, and infrastructure efficiency determine whether the system succeeds.
For many data centers, substations, ports, water facilities, and other high-consequence sites, that requirement tends to favor purpose-built thermal detection. When combined with geospatial registration, real-world range specifications, real-world AI training, and automatic PTZ control, it gives operators an early, verified alert with location, classification, and context — in the conditions that matter most.
What This Looks Like in Practice
The infrastructure question at large utility sites often proves as significant as the detection question: how many devices does the perimeter actually require, and what does that mean for total cost? At one major North American utility protecting multiple Tier 1 and Tier 2 electrical substations, because each thermal AI sensor covers detection distances and areas larger than many conventional camera designs, the utility was able to reduce the number of poles, trench runs, and power connections required across each site by about one-third compared to the alternatives considered — before any other cost factor was counted. That infrastructure reduction translated directly into lower installation cost, fewer maintenance points, and a simpler long-term support model. The selection was made on detection reliability and total cost of ownership, not unit price.
SightLogix has spent more than 20 years designing thermal AI perimeter systems for this application, including substations, data centers, seaports, airports, oil and gas sites, and government installations across North America and internationally.

PIDS Technology Selection Guide | © 2026 sightlogix.com

35


Schedule a Complimentary Consultation
Every site has a unique combination of perimeter geometry, terrain, lighting, obstructions, environmental conditions, and threat profile. A spec sheet cannot tell you how many devices your site needs, where blind spots may appear, or what detection coverage will look like in practice.
See What This Looks Like on Your Perimeter
Performance you can trust. Protection you can measure.
Made in the USA | NDAA Compliant | sightlogix.com | +1 609.951.0008 | info@sightlogix.com
A Technology Selection Guide

6
Appendix of Resources

Red Flags in AI Claims: What to Watch For
When evaluating AI-powered perimeter systems, the following are warning signs that the claim may not hold up in the field.
1
No Architecture Explanation
Vendor says "AI-powered" but cannot explain the detection architecture.
2
AI as Broad Scene Scanner
AI is used as a broad scene scanner without geospatial filtering upstream.
3
Undisclosed Training Data
Training data is not disclosed — the vendor cannot say whether AI was trained on real-world outdoor imagery or synthetic data.
4
Demo-Only Validation
Vendor relies on controlled demonstrations rather than independent field evaluations by the customer.
5
No Environmental Handling Explanation
The system cannot explain how it handles animals, vegetation, headlights, shadows, and seasonal weather change.

PIDS Technology Selection Guide | © 2026 sightlogix.com

38

RFP and Site Acceptance Proof to Add
Use the questions above in conversation. Then put the proof into the RFP and the site acceptance test, so every vendor is answering the same field problem.
A Strong RFP Should Ask For:
  • Target type, approach geometry, required range, alarm latency, confidence threshold, and final installed configuration
  • Expected nuisance alarm rate by zone, with comparable deployment evidence and the environmental causes the vendor expects to see
  • Target location precision, PTZ or confirmation-camera handoff, VMS/PSIM/access-control event fields, and behavior during network loss or server outage
  • Five-year cost per protected meter, including devices, poles, foundations, trenching, conduit, power, network, licenses, PTZ/verification cameras, lighting or IR where required, maintenance, spares, and nuisance-alarm response overhead
A Practical Site Acceptance Test Should Include:
  • A configuration freeze before testing: device locations, fields of view, firmware/software versions, analytics settings, zones, PTZ presets, integrations, and network topology
  • Human inbound, parallel, diagonal, stop-start, and low-and-slow scenarios, with vehicle and authorized-movement scenarios where relevant
  • Nuisance drivers where safe: vegetation, animals, weather, headlights, reflections, site traffic, and seasonal conditions
  • Measurement of detection, classification, alarm latency, target location precision, PTZ handoff time, operator workflow, failure behavior, and nuisance alarm causes
  • A 30/60/90-day alarm review after commissioning, with retesting after tuning changes that affect required detection scenarios

PIDS Technology Selection Guide | © 2026 sightlogix.com

39

Questions to Ask Before Procurement
Use these questions to structure vendor conversations, internal evaluations, and RFP requirements before a system is specified or purchased.
01
Primary Detection Zone and Required Warning Distance
What is the primary detection zone — and what distance from the boundary does the site require early warning?
02
Required Response Time
How much response time is required between detection and a security team reaching the location?
03
False Alarm Tolerance
What false alarm rate can operators realistically tolerate before alert fatigue affects response behavior?
04
Environmental Conditions
What environmental conditions must the system perform through — rain, fog, snow, wind, glare, darkness, vegetation, animals?
05
Target Classification Requirements
What target types must be reliably classified — persons, vehicles, animals — and at what distances?
06
Alert Response Workflow
What happens after an alert is generated — who receives it, how, and what action is expected?
07
Target Location Communication
How is the target location communicated — vague zone, camera view, or GPS coordinates?
08
Five-Year Cost Per Meter
What is the five-year cost per meter protected, including all infrastructure, false alarm response overhead, and maintenance?

PIDS Technology Selection Guide | © 2026 sightlogix.com

40

Perimeter Detection Evaluation Checklist
Use this checklist in vendor meetings, RFP development, and internal evaluations. A vendor who cannot answer these questions concretely has not demonstrated that their system performs in the conditions this guide describes.

PIDS Technology Selection Guide | © 2026 sightlogix.com

41

Independent / Regulatory / Research Sources
[S1] National Protective Security Authority
Perimeter Intrusion Detection. npsa.gov.uk. Relevant for PIDS categories, alarm verification, false-alarm causes, and no-zero-false-alarms caution.
[S2] NERC CIP-014-3
Physical Security. nerc.com. Relevant for utility physical security planning requirements.
[S3–S5] Federal Aviation Administration
Part 107 Waivers, 14 CFR Part 107, and Part 107 Waiver Safety Explanation Guidelines. faa.gov. Relevant for drone waiver framing and operating constraints.
[S6] CISA CFATS Risk-Based Performance Standard 1-7
Detection and Delay. cisa.gov. Archived guidance used only for risk-based detection/delay framing.
[S7–S11] Sandia National Laboratories
Physical Security, ReKon integrated detection, deliberate motion analytics fused radar and video, nuisance alarm data system, and impact of false and nuisance alarms. sandia.gov. Relevant for performance-based physical security, layered detection, nuisance alarm characterization, and operator trust.
[S12] Sensors / MDPI
Thermal Imager Range: Predictions, Expectations, and Reality. mdpi.com. Relevant for Johnson Criteria/DRI and field range limitations.
[S13–S15] Sensors / MDPI and IntechOpen
Modeling LiDAR through adverse weather; empirical analysis of LiDAR detection performance degradation in rain and fog; perimeter intrusion detection by video surveillance survey. Relevant for LiDAR and camera adverse-condition validation and outdoor video analytics limitations.
[S16] National Safe Skies Alliance /

PIDS Technology Selection Guide | © 2026 sightlogix.com

42

Vendor and Category References
These vendor and category references offer a snapshot of the companies and technologies that contribute to the perimeter security landscape. While their stated capabilities are noted, their inclusion here is for market orientation and does not constitute independent proof of performance superior to competitors. For objective validation, refer to the independent sources and rigorous testing outlined elsewhere in this guide.
[S17] SightLogix
AI-Powered Thermal Security Cameras for Critical Infrastructure. sightlogix.com. A key player in thermal security solutions, providing advanced cameras and a proprietary design tool for robust perimeter protection.
[S18] SightLogix
SightSurvey Design Tool. sightlogix.com. Offers sophisticated tools for planning and optimizing thermal camera deployments, crucial for effective perimeter design.
[S19] Axis Communications
AXIS Perimeter Defender. axis.com. Known for its network cameras and intelligent video analytics, providing scalable solutions for perimeter intrusion detection.
[S20] Teledyne FLIR
FLIR FH-Series ID. flir.com. A leading manufacturer of thermal imaging cameras, offering high-performance solutions for surveillance and security applications.
[S21] Hanwha Vision America
PNV-A9081RX AI Camera. hanwhavisionamerica.com. Develops advanced AI-powered cameras that integrate sophisticated video analytics for enhanced detection and classification capabilities.
[S22] Bosch Security Systems
Perimeter Security Application Note. boschsecurity.com. Offers comprehensive integrated security solutions, including video surveillance, intrusion detection, and access control for various environments.
[S23] Motorola Solutions / Avigilon
Video Security & Access Control. motorolasolutions.com. Provides integrated video security and access control systems, often for large-scale enterprise and critical infrastructure deployments.
[S24] Senstar
Fence Sensors. senstar.com. A specialist in physical perimeter intrusion detection systems, particularly known for its extensive range of fence-mounted and buried cable sensors.
[S25] Southwest Microwave
Integrated Perimeter Security Solutions. southwestmicrowave.com. Offers high-security perimeter intrusion detection solutions, including microwave and taut-wire systems for critical sites.
[S26] Ava Group / Future Fibre Technologies
Future Fibre Technologies Corporate Brochure. theavagroup.com. Specializes in fiber optic sensing technology for perimeter security, providing distributed acoustic sensing (DAS) solutions.
[S27] Fiber SenSys / OPTEX
Home. fibersensys.com. Delivers fiber optic intrusion detection systems, known for their resilience and performance in challenging environmental conditions.
[S28] RBtec
IRONCLAD Fence Alarm System. rbtec.com. Provides robust and cost-effective perimeter security solutions, including fence-mounted vibration detection systems.
[S29] Sintela
Perimeter Intrusion Detection System PIDS. sintela.com. Offers comprehensive PIDS solutions, integrating various sensor technologies for tailored perimeter protection.
[S30] Magos Systems
Ground Based Radars for Enhanced Site Protection. magossystems.com. Innovates in radar-based detection systems, offering wide-area surveillance for challenging outdoor environments.
[S31] Spotter Global
Spotter Global. spotterglobal.com. Provides ground-based radar systems designed for long-range detection and surveillance, especially in remote or large perimeters.
[S32] Navtech Radar
Perimeter Intrusion Detection. navtechradar.com. Offers advanced radar solutions specifically for perimeter intrusion detection, known for performance in all weather conditions.
[S33] Ouster
Security. ouster.com. Develops high-resolution LiDAR sensors, which are increasingly utilized in security applications for precise 3D detection and mapping.
[S34] OPTEX Europe
REDSCAN Pro Series. optex-europe.com. A provider of advanced LiDAR sensors, offering precise detection zones and virtual walls for comprehensive perimeter security.
[S35] Skydio
Site Security. skydio.com. Specializes in autonomous drone technology for various applications, including site security, offering intelligent aerial surveillance.
[S36] Percepto
Autonomous Drone Tech & Industrial Solutions. percepto.co. Develops drone-in-a-box solutions for autonomous inspection and security, enabling persistent aerial monitoring.
[S37] Easy Aerial
Tethered UAS & Drone-in-a-Box Systems. easyaerial.com. Offers tethered and drone-in-a-box systems for continuous aerial surveillance and rapid deployment in security scenarios.
[S38] Sunflower Labs
Autonomous Drone Security System. sunflower-labs.com. Provides fully autonomous drone security systems, integrating AI-powered sensors and drones for proactive perimeter defense.

PIDS Technology Selection Guide | © 2026 sightlogix.com

43