Total Sensors
5
Online
0
Offline
5
Today's Visitors
7262
β‘ 1,000 virtual sensors, concurrent ingestion β load proof
See ingestion throughput and loss rate in numbers. (Measured with a self-generated signal, no external API.)
Not run yet. Click the button to measure throughput and loss for 1,000 concurrent sensors.
Visitor Trend
Offline Sensor Alerts
Manhattan 1 Entrance
ID: MS-001 |
Last: 2026-10-25T08:41:46.618747+00:00
Manhattan 1 Exit
ID: MS-002 |
Last: 2026-10-25T08:41:46.632689+00:00
Manhattan 2 Entrance
ID: MS-003 |
Last: 2026-10-25T08:41:46.646653+00:00
Brooklyn 1 Entrance
ID: FL-001 |
Last: 2026-10-25T08:41:46.660201+00:00
Brooklyn 2 Entrance
ID: FL-002 |
Last: 2026-10-25T08:41:46.674124+00:00
Overview
Foot-traffic analytics
What FootFlow is
FootFlow is a SaaS foot-traffic analytics platform that ingests and aggregates Milesight (MQTT) and FLIR (Webhook) sensor data in real time, presenting per-store and per-floor visitor counts, occupancy and trends on a single dashboard. It supports multi-site, multi-floor structures with built-in 4-tier RBAC.
What you can do
| Category | Items | Type |
|---|---|---|
| β Core | Sensor ingestion (MQTT, Webhook), live dashboard, heatmap, trends, 4-tier RBAC, multilingual (KO/EN/JA), Excel export | Core |
| β Advanced | Floor-plan view, floor analysis, visitor-type analysis, zone flow, similarity analysis, Telegram alerts | Advanced |
| β Not in this demo | Mobile app, POS integration, camera vision analysis, external ERP integration | N/A |
How data flows
Milesight
MQTT
MQTT
β
FLIR
Webhook
Webhook
β
Ingest
FastAPI
FastAPI
β
PostgreSQL
Aggregate
Aggregate
β
Dashboard<br>live
β
Excel<br>export
1-min interval | up to 1,000 sensors | auto offline detection | imputation before aggregation
Technical features
Dual protocol
MQTT + Webhook simultaneously. Vendor-agnostic integration.
Anomaly imputation
On sensor fault, auto-estimates via adjacent-zone ratio + sales cross-check.
Similar-pattern analysis
Cosine similarity finds past similar days for operational forecasting.
Multilingual SaaS
KO/EN/JA dynamic switch, persisted, extendable with one JSON file.
About the demo data
β οΈ The figures shown are example data generated by a simulator, not real sensors.
Real-device integration (Milesight, FLIR, etc.) is built on standard protocols (MQTT 3.1.1, HTTP Webhook), so it works the same once configured for on-site hardware.
Real-device integration (Milesight, FLIR, etc.) is built on standard protocols (MQTT 3.1.1, HTTP Webhook), so it works the same once configured for on-site hardware.
Built with
Tech stack
Python Β· FastAPI
MQTT (Mosquitto)
Webhook (XML/JSON parsing)
PostgreSQL Β· time-series aggregation
APScheduler batch
4-tier RBAC Β· JWT
Jinja2 SSR Β· Chart.js
i18n (KO/EN/JA)
Docker Β· Railway deploy