1.1 System Architecture

The IK10 vandal-resistant surveillance system is structured as a multi-layer architecture where each layer has defined responsibilities, data flows, and protection boundaries. The architecture is designed so that an attack on the front-end edge does not compromise the ability of the system to record, alarm, and preserve evidence. Redundancy is built in at every critical junction — from dual-uplink fiber rings at the network layer to primary/standby VMS clusters at the compute layer.

The core deployment boundary encompasses IK10 front-end cameras, protected installation hardware, PoE network infrastructure, VMS with recording, UPS-backed power, network segmentation, and alarm rules. Optional extensions include edge AI analytics boxes for advanced behavioral detection, secondary recording locations for dual-path evidence preservation, and dual-site replication for mission-critical environments.

IK10 System Components Architecture
Figure 1.1: IK10 Surveillance System Components Architecture — Module Relations, Data Flows, and Deployment Boundaries

Data and Control Flows

  • Video Stream: RTP/RTSP or vendor SDK → VMS ingest → storage write → indexing for search and playback
  • Event Flow: Tamper/IO/analytics events → VMS rule engine → operator console + linkage devices (strobe, siren, intercom, access control)
  • Management Flow: VMS/camera management via HTTPS/ONVIF over a dedicated management VLAN, isolated from camera traffic
  • Power Flow: UPS → PDU → PoE switches and servers; SPD at all outdoor entry points → grounding bus → racks and poles

Deployment Boundaries

BoundaryComponentsRationale
Core (Mandatory)IK10 cameras, anti-tamper mounts, protected cabling, PoE network, VMS, RAID storage, UPS, alarm rulesMinimum viable system for deterrence, detection, and evidence
Optional (Enhanced)Edge AI compute boxes, secondary recording site, dual-site VMS replication, acoustic sensors, anti-spray coatingsExtends capability for high-risk or mission-critical environments

1.2 Components & Functions

The system comprises seven functional groups, each with distinct responsibilities, inputs, outputs, and key performance indicators. Understanding the role and failure modes of each component is essential for designing a system that performs reliably under adversarial conditions. The diagram below provides a grouped overview of all major components.

IK10 System Components Overview
Figure 1.2: IK10 Surveillance System — Component Groups and Inventory Overview

The table below provides a detailed breakdown of each major component category, including its responsibilities, inputs, outputs, key engineering KPIs, and the most common mismatch risk observed in field deployments.

ComponentResponsibilitiesInputsOutputsKey KPIsCommon Mismatch Risk
IK10 Dome CameraClose-range identification, tamper detectionPoE power, network, lightingVideo stream, tamper eventsIK10, IP66/67, WDR ≥120 dB, min illumination, tamper alarm latency"IK10 only" but exposed cable; poor mount leads to pry-off
IK10 Bullet CameraLonger-range view, perimeter lanesPoE power, networkVideo stream, eventsIK10 or IK10 housing, long lens, stable bracketBracket not IK-rated; cable gland weak
Multi-sensor PanoramicWide coverage, reduce blind zonesPoE power, networkMulti-stream videoCoverage continuity, calibration stability, H.265 efficiencyWrong height causes face pixel density too low
Anti-tamper Mount/BaseResist pry/twist attacksStructural substrateMechanical retentionPull-out strength, corrosion resistance, torque specMounted to weak substrate (decorative panel)
Protected Junction BoxHide cable terminations; prevent accessCable feedsSealed connectionsIP66+ rating, anti-open tamper switchPlastic box cracks; no tamper switch fitted
Industrial PoE SwitchPower and network edge distributionAC/UPS input, fiber uplinkPoE ports, VLAN, eventsPoE budget, surge tolerance, temperature range, ring protocolUnder-sized PoE budget causes reboot loops
VMS Servers (HA)Stream ingest, alarm rules, user access controlVideo streams, eventsRecordings, alarms, exportsFailover time, max channels, rule engine capacitySingle point of failure; no standby configured
Storage RAID/ECRetention and recording integrityWrite IO from VMSPlayback, exportSustained write speed, rebuild time, SMART monitoringRebuild too slow; disk mismatch causes degraded performance
UPS + SPD + GroundingPower continuity and surge protectionMains AC powerConditioned, protected powerRuntime, transfer time, SPD protection classUPS not sized for PoE peak load (IR + heaters at night)
Strobe/Siren/IntercomDeterrence and response linkageAlarm trigger from VMSAudio/visual deterrence, two-way audioActivation latency, sound level (dB), IP ratingNot triggered by tamper events; wrong zone linkage

Mismatch Consequences (Table 2)

MismatchFailure MechanismField SymptomImpactPrevention
IK10 camera + weak bracketBracket bends; camera re-aimedView shifted; scene change alarmEvidence lost from critical angleUse IK-rated bracket + anti-rotation pins
Exposed pigtail cablesCable pull/cut in secondsStream loss; camera offlineNo recording from that cameraConcealed conduit to locked junction box
No tamper rules configuredAttack not detected by VMSSilent blind spotDelayed response; no alarmConfigure occlusion/scene-change alarms with debounce
No redundancy coverageSingle camera loss = blind zoneAttacker unseen after destructionLow deterrence; evidence gapOverlapping FoV + cross-coverage pairs
Storage without write marginBurst traffic drops framesGaps in footageEvidence gap at critical momentsSize storage IO with ≥30% headroom

1.3 Working Principle

The system operates through three distinct phases: startup, normal operation, and exception/recovery. Understanding these phases and the exception chains they trigger is critical for configuring alarm rules, testing acceptance criteria, and training operators.

Startup Sequence

The startup sequence follows a strict dependency order: UPS comes online first, ensuring power stability before any network or camera equipment boots. PoE switches then boot and begin providing power to cameras. Cameras receive power, establish network links, obtain IP addresses (DHCP or static), and synchronize time via NTP. The VMS then registers all camera streams and applies recording schedules. Finally, tamper analytics are armed and alarm rules become active.

Normal Operation

During normal operation, the system performs continuous or event-based recording depending on the risk profile of each zone. Health polling runs continuously — monitoring PoE draw, stream availability, storage write health, UPS battery status, and cabinet door events. Alarms are displayed on the operator console with priority routing. Evidence export is controlled by role-based permissions with full audit logging.

Exception Chains

Exception Chain 1 — Occlusion (tape/cloth/spray): Image histogram collapses → scene change detected → defocus increases → camera tamper event fires → VMS correlates with adjacent witness camera → immediate strobe/siren + operator pop-up + incident ticket → bookmark pre/post event clips. Recovery: onsite removal/cleaning; verify lens integrity; adjust sensitivity if false positives occur.

Exception Chain 2 — Cable Pull or Cut: PoE power drop or link down → switch port down alarm + camera offline alarm → adjacent camera continues recording → VMS flags "critical blind area" → dispatch with route guidance. Recovery: inspect conduit/junction; replace cable; verify shielding/ground; test PoE class draw.

Exception Chain 3 — Bracket Sabotage (twist/aim shift): Camera orientation changed → scene-change detected → scene-change alarm fires → adjacent camera captures suspect → optional gyro/angle sensor triggers additional alert. Recovery: re-align camera; add anti-rotation pins; torque-check all fasteners; consider higher mount or cage.

PhaseKey ActionsFailure IndicatorRecovery Mechanism
StartupUPS → PoE → Camera → NTP → VMS → Analytics armedCamera not registering; NTP driftAuto-reconnect; template re-apply
Normal OperationContinuous recording; health polling; alarm display; export controlStream loss; storage write errorAlert + auto reconnect; standby VMS takeover
Uplink FailureRing reroutes; VMS standby takes over; storage controller failoverMultiple cameras offline; ring alarmRing protocol reroutes within seconds
Camera ReplacementAuto-provision templates applied; tamper re-armedNew camera not provisionedCamera template library; DHCP reservation