How IN2SOLUTION Uses Scylla AI to Protect Greece's Forests

Public Safety Case Study

How IN2SOLUTION Uses Scylla AI to Protect Greece's Forests

A wildfire spotted early is a wildfire that can be contained. A wildfire spotted late is a headline. For IN2SOLUTION, a security systems integrator and Remote Video Monitoring Center based in Athens, early forest fire detection presents a demanding operational challenge. During Greece’s summer fire season, operators would otherwise need to monitor multiple camera feeds continuously for 10 to 12 hours during daylight, watching for the first faint signs of smoke.

IN2SOLUTION addressed this challenge with Scylla AI’s smoke and fire detection technology. During the 2026 fire season, Scylla detected developing smoke at approximately 5 kilometers and automatically alerted IN2SOLUTION’s monitoring center within seconds. An operator verified the event, authorities were notified immediately, and firefighting forces were deployed in time to contain the fire before it developed into a significantly larger incident, despite difficult weather conditions and strong winds.

Challenge

The monitored areas are mountainous and remote, often without conventional power or telecommunications infrastructure. This makes continuous manual monitoring both operationally demanding and difficult to deploy at scale.

The solution needed to provide automated detection while maintaining a clear workflow for human verification and response.

Our Solution

IN2SOLUTION deployed Scylla AI’s smoke and fire detection across seven cameras: six fixed-view cameras and one PTZ camera with 32x optical zoom for manual tracing. Different lenses and fields of view were selected to optimize detection at both shorter and longer distances, providing approximately 360-degree coverage.

The system is installed on a portable, fully autonomous structure designed for seasonal deployment in strategic high-risk locations. Solar panels provide power where there is no grid access, while Starlink provides connectivity to IN2SOLUTION’s infrastructure. The unit is returned to IN2SOLUTION’s facilities during winter for maintenance and upgrades.

Scylla continuously analyzes every video stream and automatically generates an alarm when potential smoke is detected. The alert is sent to the Remote Video Monitoring Center, where an operator reviews live and recorded footage, verifies the event and initiates the appropriate escalation procedure.

AI performs the continuous detection work; human operators retain responsibility for verification, judgment and response.

Scylla’s smoke and fire detection is designed to reduce false positives from common visual triggers such as sunlight, clouds and nighttime lighting, and reports up to 98.5% accuracy.

Protecting the Monitoring Infrastructure

The remote monitoring units also need protection from unauthorized access, vandalism and theft. IN2SOLUTION uses Scylla’s perimeter intrusion to establish a security zone around the installation.

When a person approaches or enters the protected perimeter, an alarm is sent to the monitoring center. Operators can verify the event through live video and communicate with the individual through the unit’s loudspeakers, warning them that the area is monitored and instructing them to move away. The same AI platform therefore protects both the forest being monitored and the infrastructure doing the monitoring.

One Platform, Multiple Applications

IN2SOLUTION’s Remote Video Monitoring Center manages several thousand cameras, primarily for perimeter protection and remote surveillance. Through Scylla, the same AI ecosystem supports:

Perimeter intrusion detection

Early smoke and fire detection in forest areas

Boat and vessel detection at aquaculture facilities

All three applications can be managed within the same monitoring environment and established operational procedures, giving a monitoring center operating at scale a standardized approach to different types of AI-generated events.

Results

● Smoke detected at approximately 5 kilometers during a real-world event in the 2026 fire season ● Automated alert delivered within seconds to the monitoring center ● Immediate human verification and escalation, enabling authorities to be notified without delay ● Fire contained before becoming a significantly larger incident, despite difficult weather conditions and strong winds ● Protection of both the forest and monitoring infrastructure through the same AI platform ● Multiple AI applications consolidated within a single monitoring environment

Customer testimonial

For us, this incident demonstrated the real operational value of AI-based early smoke detection: early detection, immediate human verification, and rapid response at a stage when every minute can make a significant difference.

Stathis Bimpas

Project Manager at IN2SOLUTION