Simat Labs
Book a demo

Queues

Queues

Queue length at counters, gates, and desks from existing cameras, with alerts when lines build up.

Store 04 — Front of store

--:--

LIVE·CAM-04-02·no faces stored

In store

412

Flow

38/min

Waiting

23

checkpoints
density · last 60s
C1

Checkout 1

6 in queue

C2

Checkout 2

11 in queue

C3

Checkout 3

4 in queue

SD

Service desk

2 in queue

edge-04 · healthy

pax/min

Queues measures how many people are waiting in a line — at a checkout, a service desk, a gate, or a counter — from the cameras that already cover it.
Teams see queue length live and over time, so they can open another counter before a line gets long, not after customers complain.

Detection runs on-premise and is anonymous: people are counted inside a queue zone you draw. No faces are stored and no one is identified.When a queue passes the length you set, no-code automations send an alert, and the dashboard shows when and where lines build up across the day.

Live queue length

People waiting in each queue zone you draw on the camera view, updated continuously.

Alerts before it gets long

Set a length per queue and get an email alert when it is passed, with a cooldown so one busy minute is one alert.

When lines build up

Queue length by hour and day, per counter and site, to plan staffing around real demand.

Powered by IFA

Runs as a module on the IFA edge box beside your cameras.

About IFA

The IFA dashboard

One dashboard for every module

Every IFA module reports into the same web dashboard. A new module shows up with its charts, events, and automations the day it is deployed — nothing is coded per module.

Plant overview — today

HourDayWeek·group by: line

Events per hour

Line 3 · QA failsDock 4 · crossings
020406006:0008:0010:0012:0014:0016:00Dock 4 · crossingsLine 3 · QA fails

Events by severity

Critical6
Warning41
Info137

184 events · 2 modules · 4 cameras

Streams

  • Line 3 — labeleronline · 24 fps · 41 ms
  • Dock 4 — closeonline · 20 fps · 38 ms
  • Dock 4 — fardegraded · 11 fps · 2 reconnects
  • Line 1 — capperoffline · last seen 09:12

Automation runs

  • Label fail → email shift leadok11:42
  • 5 fails in 10 min → capture + WhatsAppok11:38
  • Dock 4 far offline → emailsuppressed11:20
  • Session closed → reportok10:57

Live camera monitoring

Every camera as a health card: online, degraded, or offline, with frame rate, latency, and uptime, and a live view from the edge box when it is streaming. Refreshes on its own.

Charts derived from events

Pick a metric, group by line, zone, severity, or product, and bucket by hour, day, or week. Line, bar, area, stacked, and pie — all built from the events a module emits, never hand-coded.

Event log and notifications

Every event, filterable by module, type, and severity, with the frame that triggered it. Email and WhatsApp notifications go out through automations, with a per-run record of what was sent and what was suppressed.

No-code automations

A drag-and-connect flow builder: a trigger (a module event, a camera going offline, a schedule), logic (filter, window, cooldown, time gate), then actions — send an email, send a WhatsApp message, capture the frame, generate a report.

Reports for managers

A report action turns a session or a shift into a readable document — totals, shares, per-product detail — attached to the notification email and kept in the dashboard's report list.

Queue questions

How is a queue measured?

You draw a queue zone on the camera view. The system counts the people standing inside it and reports that as the queue length, continuously.

What cameras do we need?

A camera that sees the whole queue, usually existing CCTV with an RTSP/ONVIF stream. Long or winding lines may need a second view. Processing runs on the IFA edge box on site.

Is anyone identified?

No. People are counted inside the zone without storing faces or identifying anyone. Only queue lengths and alerts leave the site.

Who gets alerted, and how?

Anyone you choose. Alerts are no-code automations in the dashboard: pick the queue, the length, and a cooldown, then the email recipients. Every alert is logged.

Last updated: