The Greek pilot of the TREEADS project was held on October 24th and 25th, 2024, at the Samaria Gorge in Crete, Greece. This pilot addressed critical challenges, including wildfire detection and situational awareness. It also focused on improving response coordination in steep and inaccessible terrains.
Day 1: Presentations and Preparations
On October 24th, the activities commenced at MAICh with a series of structured presentations. These sessions introduced stakeholders to the technologies and tools to be demonstrated during the pilot.
The day began with an overview of TREEADS objectives, setting the stage for discussions on fire prediction, monitoring, and response. The Fire Weather Index, presented by the National Observatory of Athens (NOA), showcased a high-resolution model to predict wildfire risks. Following this, NCSRD presented models for fire plume dispersion and evacuation scenarios, showcasing tools that predict fire spread and crowd movement during emergencies.
Further, emphasis was placed on multi-sensor technologies and evacuation support systems, which provide real-time environmental data to responders. ACCELIGENCE demonstrated their UAV-based surveillance systems, which specialize in wildfire detection in complex terrains. CERTH presented their AI-powered mission planning software that optimizes drone flight paths for efficient monitoring.
A highlight of the session was the introduction of AR Helmets developed by 8BELLS. These helmets, used by responders during the simulation, integrate environmental sensors, the X/BELLO communication system, and Black Box units. They provide navigation assistance, situational awareness, and live communication, enabling safer and more coordinated operations. Additional tools included the Event-Driven Situation Detector and the Wildfire Response Engine by FRONTIERS INNOVATION, which support dynamic decision-making during fire events. Finally, EFΒ presented their Open Street Map-based mapping tools, providing enhanced geographical visualization capabilities.
Further technologies demonstrated included to the Air quality multi sensor unit (MSU) and evacuation application for mobile phones by NCSRD, the social network analysis of the Samaria Gorge stakeholders (CBS) and finally the Economic risk assessment from wildfires in Crete (DTU)
Day 2: Live Demonstration at Samaria Gorge
On October 25th, participants gathered at the Ksiloskalo region, the entry point to the Samaria Gorge, for hands-on field demonstrations. The demonstration focused on the integration of TREEADS technologies. ACCELIGENCE launched drones equipped with thermal imaging cameras to detect and monitor fire hotspots in real time. The thermal and visual data, processed onboard using the NVIDIA Jetson AGX Xavier, were streamed live to the Command-and-Control Center (CCC), allowing decision-makers to allocate resources quickly and effectively.
Another key activity was the demonstration of a full simulation scenario of a wildfire in the Gorge, executed by FRONTIERS INOVATION and EFB, (incorporating information flow from CERTH/8 BELLS/ACCELI/NCSRD/USAL/CAPGEMINI/MAG). This full scenario demonstrated the ingestion of data from diverse sources (such as environmental IoT sensors or images stemming from UAVs), and the detection critical situations leveraging knowledge elicited from the main crisis management actors. During the full scenario, the system recognized and interchanged Five Stages of Emergency based on the different State Transition Triggers as dictated from the environmental conditions sensors. Furthermore, the system utilized data from the fire and smoke propagation models, as well as the iCrowd simulator. The visualization of the data streams and the emergency stages was performed on the OSM platform (EFB).
The iCrowd simulator was also showcased during the demonstration. This tool modeled crowd movement patterns during evacuation scenarios, helping stakeholders analyze and refine their response protocols. Additionally, plume dispersion models predicted smoke spread, further enhancing situational awareness.
Key Technologies Demonstrated
CERBERUS UAV Deployment
The CERBERUS UAV, an advanced octa-rotor platform developed by ACCELIGENCE, played a critical role during the pilot. Designed for wildfire detection and response, the UAV combines precision navigation, extended endurance, and integrated sensor capabilities. Equipped with the FLIR DUO PRO-R sensor, CERBERUS captured high-quality thermal and RGB imagery, enabling precise detection of fire hotspots. The onboard object detection module identified vehicles, humans, and animals; streaming data live to the ground control station for situational analysis.

Visual Object Detection results from the Greek Pilot test illustrating detected vehicles

Visual Object Detection results from the Greek Pilot test illustrating detected people and vehicles
Key highlights of CERBERUS included its real-time data transmission capabilities, precision hotspot detection, and rapid deployment for dynamic wildfire scenarios. Challenges such as abrupt drone maneuvers and strong winds were addressed by refining flight strategies and improving video resolution to align with detector training standards.
Hotspot Detection System
The hotspot detection system, calibrated by ACCELIGENCE, was rigorously tested during the Greek pilot. It demonstrated high accuracy in detecting residual heat sources in extinguished fire zones, even in the complex terrain of the Samaria Gorge. The system dynamically created UAV flight missions and synchronized operations with TREEADS tools, such as the WebGIS and AR/VR systems, providing real-time situational awareness.

Integration of Hotspot Detection of the UAV with a map indicating the fire and the location of the UAV

Hotspot Detection results from the Greek Pilot test illustrating detected fire
Simulation Tools and iCrowd Integration
The iCrowd tool, along with the evacuation simulations, was showcased to participants as part of the live demonstrations. A 3D model of the Samaria Gorge was used to predict and visualize evacuation routes, allowing stakeholders to assess crowd flow and warden-agent coordination. This tool significantly contributed to refining evacuation protocols in high-risk wildfire scenarios.

Evacuation simulation, training and decision support: crowd being redirected by warden agents

Samaria Gorge (Greek Pilot Site area) 3D model on iCrowd Simulator
Demonstration of the Air quality multi sensor unit and evacuation application for mobile phones
Another demonstrated technology was the NCSRD-MSU, which is a portable IoT device designed for real-time monitoring of air quality characteristics & pollutants such as Temperature, Humidity, Carbon Monoxide (CO) concentration and Particulate Matter (PM) in various fractions of mass and number concentrations. This device was developed to enhance first responders’ (FR) safety situational awareness and interconnectivity between the FR, their wearables, and command and control centers.
Additionally, the Evacuation Support Mobile App was demonstrated. This app has two modes, a first responder mode and a citizen mode. The app features a live map that allows oversight of first responder status, enhanced with real-time sensor readings from NCSRD-MSU devices carried by responders. Additionally, features notifications (in both modes) with added particle hazard alerts for abnormal sensor values, along with a First Responder status screen, analytics, and detailed charts.

The MSU (left) and snapshots of the Evacuation Support Mobile App (center and right)
Outcomes and Lessons Learned
The Greek pilot successfully demonstrated a holistic and technology-driven approach to wildfire management. By integrating UAV-based surveillance, AI-powered tools, AR-enabled navigation systems, and crowd simulation models, TREEADS showcased advancements in situational awareness, response coordination, and decision-making.
The CERBERUS UAV and hotspot detection systems proved highly effective in identifying and monitoring fire hotspots, while AR helmets and the X/BELLO system enhanced responder safety and communication. The pilot also highlighted areas for improvement, such as optimizing drone manoeuvres, increasing video resolution, and aligning detector parameters with field conditions.
The full scenario demonstrated the complementarity of the TREEADS technologies and the added value in the decision making and response time.
Overall, the Greek pilot validated TREEADS technologies in a challenging real-world environment, emphasizing their potential to revolutionize wildfire management. The outcomes reinforced the importance of integrating science-based tools to protect human lives and natural ecosystems while addressing the unique challenges of wildfire-prone regions.