Solutions / Camera-based fire detection
Let your cameras see fire.
AICO Vision uses AI to analyse the footage from your facility’s existing security cameras in real time. The moment a flame or smoke appears, it alerts you with a snapshot showing which camera and where — day or night.
- Training images
- 350,000+
- Day and night
- 24/7
- Cameras in field testing
- 26
- New cameras needed
- 0
You do not need new cameras. Thanks to an AI trained on more than 350,000 images, your existing cameras start to “see” flames and smoke. You find out while the fire is still small, before the smoke even reaches the ceiling detector.
Live simulation
Day or night, 6 cameras on one screen.
Watch 6 cameras of a facility live. Start a flame or smoke on a camera and see how many seconds AICO Vision needs to catch it. Try misleading images like sun glare, welding light or headlights; switch to night mode and see it work on infrared video too.
Scenes and detection are simulated for this page; in the real system your cameras’ live video is analysed by an AI model trained on more than 350,000 images. In night mode cameras switch to infrared (IR): flame appears very bright and smoke as a grey haze. An alarm is raised when the probability stays above 60% for a set time.
Images the AI was trained on
Why it matters
Conventional detectors wait for smoke to reach the ceiling.
In high-ceiling warehouses, large production halls, car parks and open areas, smoke can take minutes to reach a detector — and in open areas it never does. Yet most of these areas are already watched by cameras 24/7; nobody can simply watch every screen at once.
How the AI was trained
It saw hundreds of thousands of examples to learn what fire looks like.
AICO Vision’s AI model was trained on more than 350,000 images. They cover flames of different sizes, thin and dense smoke, day and night (infrared) conditions, and indoor and outdoor spaces. Just as important are images that look like fire but are not: because the model has seen these too, it raises fewer false alarms.
Flames of different sizes, colours and distances — from small sources like candles to large fires.
Thin, dense, white, grey and black smoke against different lighting and backgrounds.
Night camera footage with infrared lighting, low light and noisy frames.
Sun glare, headlights, welding light, orange workwear, steam, fog and dust.
How it works
Three simple steps.
- Step 1
Watches
Footage from your existing cameras is analysed continuously, dozens of frames per second.
- Step 2
Tells apart
The AI recognises flames and smoke and filters out misleading images such as sun glare, headlights and orange workwear.
- Step 3
Alerts
Which camera, which area, what time — you get a notification with a snapshot through the channel you choose.
Features
Every detail considered.
Flames and smoke, together
Flames and smoke are analysed simultaneously in the same video stream, with a separate probability for each.
Day and night
Works on colour footage by day and on the camera’s infrared (IR) footage at night. In the dark, flames appear very bright and smoke shows up as a distinct haze.
Works with existing cameras
Connects to your compatible IP camera streams; no new cameras or cabling required.
Fewer false alarms
Misleading images such as sun glare, headlights, welding light, orange clothing and steam are part of the training data.
Works even if the internet goes down
Analysis continues locally; events are recorded and sent once the connection is back.
Separate view for each camera
Flame and smoke probabilities, event counts and times for each camera on a single screen.
Notifications with snapshots
Each notification includes a snapshot of the event with the area marked, so the team knows what to expect.
The notification channel that suits you
Mobile app, SMS, email or your existing alarm panel — the method is chosen per project.
Developed in the field
Tested and refined in a real 26-camera building with controlled flame and smoke scenarios.
Who is it for?
Where AICO Vision makes the biggest difference.
- Warehouses and logistics hubs
- Production plants and factories
- Car parks and open sites
- Shopping malls, hotels, hospitals and public buildings
- Energy and infrastructure facilities
- Historic buildings and museums
Technical specificationsFor engineers and technical teams
| Detection | Flames and smoke, simultaneously |
|---|---|
| Operation | 24/7 · day and night (IR) footage |
| Training data | 350,000+ images |
| Cameras | Existing, compatible IP cameras |
| Connection loss | Analysis continues locally, events are stored |
| Notification | Per project: app, SMS, email, alarm panel |
| Field testing | Active field study with 26 cameras |
| Monitoring interface | Live view and probabilities per camera |
Performance is assessed based on the project site, camera placement, image quality and ambient conditions. AICO Vision is positioned as an early detection layer that complements mandatory fire detection systems, not as a replacement for them.
Try it free with your cameras.
Contact us for a free technical consultation and demo. We reply within one business day.
Other solutions
You might also like.
AICO Sense
Catches the very first gases of smouldering, before heat builds up or smoke appears.
VibrationGuard
Tells you a machine will fail two weeks in advance.
ColdTrack
Tracks the temperature of medicines and food at every moment of the journey.