Guard Watcher Agent
Confirms guards are present at post and not sleeping on duty — with two-way audio wake-up, supervisor SMS escalation, and a daily compliance scorecard.
“Who is watching whether the guard is actually watching?”The moment a post goes unmanned, you know.
Everyone learns to live with the gap between “something happened” and “someone noticed.” We did not. Let the cameras do the watching and your people do the deciding — alerted the instant it matters, not hours later. From “that is just how it works” to “wait — it could work like this.”
You pay for a guard. You are paying for whatever the guard does at 3 a.m.
Every operations head has the same story. The CCTV footage from last night shows the guard slumped on a chair for hours. The patrol log claims he walked the perimeter every 60 minutes. The breach you are investigating now occurred at 03:14 — exactly when the recording shows nobody was awake.
This is not a guard problem. This is a system problem. Manual patrol logs are signed in advance. Punch-clock systems are punched by the next guy. Senior management discovers the gaps when something goes wrong — long after the policy is supposed to have prevented it.
And nobody is watching the watcher. The supervisor cannot stay awake for the night shift. The CCTV is reviewed after the fact. You have cameras pointed at the inside of the building, but none of them are actively watching whether the security layer is doing its job.
The watchdog with no shift change.
Guard Watcher Agent monitors every guard post continuously. Per the VIZO361 datasheet, it extends Use Case 17 (Sleeping on Duty Detection) with patrol verification across the existing camera footprint. It confirms each guard is present, alert, and at the right post during every required interval. Inactivity beyond a configurable threshold triggers a two-way audio wake-up via the camera speaker, followed by SMS / WhatsApp escalation to the supervisor.
Patrol compliance is verified the same way. Each patrol route is defined as a sequence of zones with cadence and dwell. The agent confirms each leg of the patrol — physically — using cameras already covering the route. Skipped legs are flagged. Patrolled-too-fast events (suggesting a fake walk-through) are flagged. The agent does not care what the paper log says.
A daily compliance scorecard is generated automatically — per guard, per shift, per post — and emailed to operations every morning. After 90 days you have an evidence-grade record of what your security layer actually did, not what the contractor claimed.
Known limitations (from the VIZO361 datasheet): it is difficult to distinguish between intentional rest, authorised break, and fatigue-related inactivity. The agent surfaces patterns for human review rather than issuing disciplinary decisions automatically. Acceptable inactivity windows must be configured per shift and per post.
How Guard Watcher Agent works — end-to-end
From the moment a frame leaves your camera to the moment your hardware reacts. No black box — every step explained.
Define posts + patrol routes
Each guard post tagged on the camera footprint. Patrol routes defined as zone sequences with cadence and dwell limits.
Presence + posture detection
Vision model detects the guard at post and classifies posture (standing, seated alert, slumped, head-down, absent).
Activity scoring
Continuous activity score from body keypoints and head motion. Inactivity beyond threshold triggers a concern state.
Hardware actuation
Two-way audio wake-up via camera speaker. Escalation to supervisor SMS / WhatsApp if no response. Optional shift-supervisor PA broadcast.
Daily scorecard
Per-guard / per-shift compliance score: presence percentage, patrol completion, response time to wake-up. Emailed to operations every morning.
Detection without action is just an alarm. We close the loop.
Every Guard Watcher Agent event can trigger your existing on-site hardware. Wired or wireless. PLC, contactor, BACnet, MQTT or webhook — your choice.
Two-Way Audio Wake-Up
Integration
Audio-out on camera or paired PA speaker. Multilingual pre-recorded prompts (EN / HI / AR / regional).
Outcome
Sleeping guard hears name + post + 'please respond'. Most respond quickly. No human intervention needed for routine wake-ups.
Supervisor SMS / WhatsApp
Integration
Twilio, MSG91, WhatsApp Business API, Slack, Teams.
Outcome
If guard does not respond within configured window, message + clip pushed to shift supervisor with location, name, last-known activity.
PA System Escalation
Integration
Audio output to facility PA controller for zone or facility-wide broadcast.
Outcome
Persistent inactivity escalates to a zoned PA broadcast — reserved for genuine no-response scenarios.
Daily Scorecard Email
Integration
SMTP / Sendgrid / SES. Configurable recipients per site.
Outcome
Operations head receives per-shift compliance summary every morning. PDF + CSV attached for archive.
Guard-Service Contract Reporting
Integration
Monthly export to PSP / PSA contract dashboard.
Outcome
Hard data for SLA enforcement against guarding contractor. Disputed events resolved on video evidence, not memory.
Where Guard Watcher Agent is deployed
Specific industries, specific scenarios, specific outcomes — not generic marketing claims.
Warehousing & Logistics
Yard gates, dock-doors, perimeter towers
Wide perimeter at night, single-guard coverage. Guard Watcher confirms presence per cadence.
Manufacturing
Material gate, production-floor patrol
Production-floor patrols often skipped on night shift because nothing is moving. Guard Watcher confirms physical patrol per cadence. Statutory factory-inspectorate requirements met with auditable evidence.
Residential Complexes & Gated Communities
Tower lobby, podium, parking patrols
Resident complaints about guard absenteeism resolved with hard data. Patrol cadence enforced. Building management gets visibility on actual service vs. contracted service.
Oil, Gas & Critical Infrastructure
Tank-farm towers, perimeter watch
Statutory critical-infrastructure obligations. Patrol verification with audit-grade log. Multi-shift handoff confirmed by camera.
Corporate Campuses & Office Parks
Reception, secondary gates, basement patrol
Reception-desk coverage during business hours verified. Night-shift parking patrol verified. Multi-tenant scenarios with shared security service especially benefit.
BFSI / ATM Cabins
Branch + ATM cabin guard monitoring
RBI / SAMA / regulator-mandated guard presence. Compliance auditable by camera evidence, not signed log book.
Technical specifications
The numbers we publish in proposals. Detailed accuracy targets, latency budgets, and integration scope.
| Datasheet reference | Use Case 17 — Sleeping on Duty Detection (extended with patrol verification) |
| Deployment timeline | Approximately 3 days per post, per the datasheet |
| Minimum camera resolution | 2 MP |
| Training required | Minimal — inactivity-threshold tuning during the pilot |
| Camera positioning | Existing cameras at posts and control rooms typically work; 1080p+ recommended for clear posture classification |
| Inactivity threshold | Configurable per post and per shift |
| Multilingual prompts | EN / HI / AR / TA / TE / BN / regional |
| Compatible cameras | Standard RTSP / ONVIF IP cameras, 2 MP–8 MP |
| Daily scorecard | PDF + CSV. Per-guard / per-shift / per-post. Retained for audit. |
| Privacy | Body-pose model only — no facial recognition required for Guard Watcher (optional add-on via Identity Agent) |
| Benchmark accuracy | Site-specific — measured and documented during the 30-day pilot |
Make sure you are getting what you pay for. Disputes with security contractors end with evidence.
Most enterprises already pay for 24×7 manned guarding. Adding Guard Watcher does not replace the guard — it ensures you receive the contracted service. Customers report a material drop in 'guard not at post' incidents once the team knows the system is live; the specific reduction at your site is benchmarked during the pilot.
Insurance posture: insurers underwriting industrial, commercial and residential risks increasingly look favourably on auditable surveillance + compliance evidence. Premium impact depends on jurisdiction, asset class and broker — we provide the audit pack.
SLA disputes: with third-party security service providers, the Guard Watcher evidence stream supports renegotiation when the contracted service does not match delivery. The size of any specific recovery is determined by your actual SLA terms — benchmarked during the pilot.
A real-world example
Customer name withheld per NDA. Full reference available on request.
Reference customer (named under NDA)
Proeffico has deployed Guard Watcher across reference customer sites in 3PL warehousing, industrial sites and residential complexes. Verified outcomes are available under NDA on request. Site-specific reductions in 'guard not at post' incidents and SLA-evidence recoveries are benchmarked during your 30-day pilot and shared in a signed pilot report.
Pilot
Real numbers benchmarked on your site, in your signed pilot report
Guard Watcher Agent vs the traditional way
Most facilities still run on the old playbook. Here's what changes.
The traditional way
- ✗Paper patrol log — signed in advance, no real verification
- ✗Sleeping guard discovered next morning during CCTV review
- ✗Disputes with security contractor end in he-said / she-said
- ✗Supervisor needs to physically walk all posts to verify
- ✗Insurance audit relies on paper trails and verbal assurances
With Guard Watcher Agent
- Physical patrol verified by camera, leg by leg. Skipped legs flagged in real time.
- Inactivity detected during the shift. Two-way audio wake-up. Supervisor SMS if no response.
- Auditable evidence stream. Per-guard / per-shift / per-post scorecard.
- One supervisor monitors compliance across all sites from a phone or dashboard.
- Hard evidence stream that satisfies insurer / regulator / inspectorate requirements.
Questions buyers ask before they sign
Doesn't this damage trust with our security team?
+
Most customers find the opposite. Good guards welcome it because it ends the constant suspicion. Underperforming guards self-select out. The team that remains is materially better. We recommend a 2-week 'live without enforcement' onboarding for change management.
What about toilet breaks, lunch, and legitimate absences?
+
Each post has a defined inactivity threshold and acceptable absence windows. Toilet / meal breaks are policy — configured per site, per shift. Guards can send a short PA / WhatsApp acknowledgement when stepping away. The agent learns the legitimate patterns and only flags genuine outliers.
Does it use facial recognition?
+
Not by default. Guard Watcher uses body-pose detection only. If you want per-guard attribution, you can pair with Identity Agent. This is opt-in.
What languages does the wake-up audio support?
+
English, Hindi, Arabic, Tamil, Telugu, Bengali, Marathi, Nepali, Oman-dialect Arabic, and others. Custom phrases supported.
How do we measure ROI?
+
Three dimensions, each benchmarked at your site: (1) reduction in 'guard not at post' incidents, (2) insurance posture / premium impact (jurisdiction-dependent), (3) recovery of billed-but-not-delivered guarding hours on third-party contracts.
Can the guards see when they're being watched?
+
Up to you. Most customers display a small indicator at each post for transparency. Others keep it discreet for a clinical baseline.
What if my cameras aren't pointed at the guard posts?
+
Existing cameras typically work because guard posts are usually inside camera coverage by design. Where a post is in a blind spot, a single supplementary camera per post is recommended.
What does it cost?
+
VIZO361 is sold on a Private Offer model. Pricing depends on number of posts, cameras involved, integration scope and sites. A 30-day pilot on 3–5 posts at one site is included.
From the live VIZO361° platform — not mockups
Screenshots and alert clips captured from the deployed product. Customer identifiers have been blurred where present.
Who built this — and who reviewed it
Last reviewed: 2026-05-22Product Lead
Utkarsh Yadav
AI / ML Lead, VIZO361° (Facial Recognition + Liveness Detection)
Leads the facial-recognition, liveness-detection, and identity workflows in VIZO361. Designs the agent architectures behind Identity, Footfall Analyst, and PPE Compliance.
Product Lead
Swastik Vaish
AI Engineer, VIZO361° (Fire & Smoke, Cash Detection)
Leads the safety + loss-prevention model family in VIZO361 — including Fire Watch, Cash Counter, Guard Watcher, and Crowd Density. Owns model accuracy benchmarking during pilots.
Reviewed By
Saurabh Agarwal
Founder & CEO, Proeffico Solutions Pvt Ltd
Founded Proeffico in 2018. Reviews every product narrative for accuracy against the VIZO361 datasheet, customer reference behaviour, and Proeffico's published trust commitments.
Profile →All capability claims on this page are anchored to the VIZO361 Product Datasheet (2025). Site-specific accuracy, latency, and ROI figures are benchmarked during a 30-day pilot on your cameras and shared in a signed pilot report — not published as universal numbers.
Explore other Vision Agents
Private Offer · No published list price
See Guard Watcher Agent on your cameras. 30-day pilot. NDA. Real conditions.
Two-camera pilot on your site. Real models. Real lighting. Real workflows. Quote within 5 working days of discovery.