Wildfire early warning system with “SmartForestFire”

When it comes to wildfire early detection, minutes often make all the difference. The “SmartForestFire” project is developing an AI-powered wildfire early warning system for automated wildfire detection. Cameras and AI models detect smoke at an early stage. Sensor technology provides additional measurements for situation assessment, while mioty® supports data transmission and positioning. The project examines how emergency responders can be alerted in a way that integrates seamlessly into existing operational processes. The Fraunhofer Institutes for Integrated Circuits IIS and for Material Flow and Logistics IML are testing the solution in the Tennenloher Forest near Erlangen. The goal is to detect wildfires faster, reduce false alarms and enable cost-effective monitoring of smaller forest areas.

“SmartForestFire” detects wildfires early, reliably and cost-effectively. A matter of minutes can determine whether initial smoke development escalates into a major wildfire. Rising temperatures and prolonged dry spells increase the risk. According to the World Wide Fund for Nature (WWF), wildfires caused approximately 2 billion euros in damage and destroyed 1.4 million hectares of forest in 2025. In Germany, authorities recently recorded more than 1,100 wildfires annually, destroying around 840 hectares. Every firefighting operation is costly and places a strain on emergency response organizations, many of which rely on volunteers. Smaller or difficult-to-access areas, such as the munitions-contaminated Tennenloher Forest in Erlangen, have been difficult to monitor efficiently until now. The project reduces false alarms, accelerates detection and makes monitoring of high-risk areas more affordable. The goal is to provide authorities and emergency responders with a reliable target process instead of isolated alerts.

In the “SmartForestFire” project, Fraunhofer IML connects the technology of the early warning system with the real operational processes of emergency responders. Fraunhofer IML documents and analyzes the complete operational process, from the reporting of a wildfire to the integrated control center, through firefighting activities and the completion of the operation. The analysis is based on expert discussions with the participating user organizations: the Erlangen-Höchstadt District Fire Department, the Erlangen Office for Fire and Disaster Protection and the Nuremberg Integrated Control Center. The results are consolidated into a cross-organizational current-state process model. In addition, future integration options for the “SmartForestFire-System” within existing operational processes are developed as future-state processes. Fraunhofer IML’s work includes the following steps:

• Documenting and analyzing operational processes through expert interviews with user organizations.
• Creating a cross-organizational current-state process model to visualize the findings.
• Developing integration options for the SmartForestFire system within existing operational processes.
• Designing and visualizing future integration concepts as future-state processes.
• Deriving future process changes, including newly introduced or obsolete process steps.
• Identifying optimization potential enabled by the SmartForestFire system.

Are you responsible for fire protection or disaster management? Within the “SmartForestFire” project, a demonstration system is being tested with the following functions:

• Image and video sequences from different time horizons, perspectives and zoom levels
• Integrated analyses and operational data on situational status, weather conditions and accessibility
• Indications for assessing risk situations and prioritizing further measures

Contact us

“Reliable wildfire early detection cannot be achieved through a single technology alone. Only the combination of AI-based image analysis, additional sensor technology, robust wireless communication and their integration into existing operational processes of authorities and organizations with security responsibilities (BOS), together with the experience and expertise of the emergency responders involved, creates a system that performs reliably under real-world conditions.”
Holger Schulz, M. Systems Eng., conducts research on wildfire early detection at Fraunhofer IML

“SmartForestFire” combines four components into a cost-effective solution. Cameras provide image data that specifically trained, multi-stage artificial intelligence (AI) models analyze. These models detect smoke at an early stage and reliably distinguish it from clouds, fog and dust. The linkage of sensor, image and geospatial data significantly reduces false alarms. At the same time, modern image recognition enables the use of more cost-effective hardware, an advantage for smaller high-risk areas.

The mioty® wireless technology transmits sensor data such as temperature, wind, soil moisture, or smoke in an energy-autonomous and interference-resistant manner. It functions even if infrastructure fails due to a fire. Three base stations in Marloffstein, Tennenlohe, and southern Erlangen cover an area of approximately 200 square kilometers. The system determines the sensor locations using time-of-flight measurement. All data converges on a central platform. This provides control centers with an up-to-date overview of the situation – including image sequences, analyses of forest conditions, weather and accessibility, as well as guidance on prioritization.

Within the “SmartForestFire” project, Fraunhofer IML analyzes how wildfire alerts and wildfire response operations are currently conducted and how they can be digitally supported in the future. Based on expert interviews with the Erlangen-Höchstadt District Fire Inspection Office, the Erlangen Office for Fire and Disaster Protection and the Nuremberg Integrated Control Center, Fraunhofer IML documents real operational processes and visualizes them as current-state processes. On this basis, Fraunhofer IML develops future-state processes and identifies concrete integration options for the “SmartForestFire” system.

Would you like future-proof control center operations and emergency response processes? Fraunhofer IML contributes expertise in civil security research, process analysis and digital operational process management.

Ready to future-proof your control centers or operational workflows? Fraunhofer IML contributes expertise in civil security research, process analysis, and digital operations and workflow management.

Contact us about digitizing operational processes

Parts of Franconia are among the regions in Bavaria with the highest wildfire risk. Dry coniferous forests, difficult terrain and unexploded ordnance create specific challenges. The munitions-contaminated Tennenloher Forest in the Erlangen-Höchstadt district clearly illustrates the situation. Because Fraunhofer IIS is headquartered in the Tennenlohe district of Erlangen, the forest was selected as a pilot area. Since September 2025, the first camera has been operating in test mode on one of the institute’s radio towers. The system anonymizes image data directly on the camera before it leaves the device, meaning that no personal data is generated. By late summer 2026, the team will add two additional camera locations as well as further mioty® base stations and sensors. This enables large-scale monitoring of forest conditions. Initial tests confirm the reliable detection of smoke during its earliest stages, providing a crucial time advantage. Following the internal testing phase, the system will automatically notify the Nuremberg Integrated Control Center of potential wildfires. The established infrastructure also enables new AI models to be tested under real operating conditions.

A wildfire early warning system only creates value if critical information can be transmitted reliably during an emergency. This is precisely where mioty® wireless technology comes into play. The patented Telegram Splitting technology makes data transmission robust, interference-resistant and energy-autonomous. This significantly improves resilience. Even if existing infrastructure is compromised by a wildfire, mioty® continues to transmit sensor data over long distances and alongside other wireless systems. Three base stations in Marloffstein, Tennenlohe and southern Erlangen cover approximately 200 square kilometers. Time-difference-of-arrival measurement enables precise sensor positioning. When a sensor reports an event, the development and spread of the situation can be tracked. For digital operational process management, “SmartForestFire” consolidates these data streams on a central platform.

Fraunhofer IML is evaluating how the system can be integrated into existing operational processes and identifies optimization opportunities. In this way, technology and process expertise combine to create a reliable Common Operational Picture (COP) for control centers as well as authorities and organizations with security responsibilities (BOS).

“SmartForestFire” is a pilot project of the Fraunhofer Cluster for Cognitive Technologies (CCIT). Fraunhofer IIS and Fraunhofer IML collaborate as research partners. Practical expertise is provided by the Erlangen-Höchstadt District Fire Inspection Office, the Office for Fire and Disaster Protection of the City of Erlangen, the Nuremberg Integrated Control Center, the Federal Forest Service and the Bavarian State Forests. The two camera locations outside Fraunhofer IIS are facilitated by the “Zweckverband zur Wasserversorgung der Marloffsteiner Gruppe“, Erlangen Public Utilities and the Free State of Bavaria, which are providing suitable sites. The solution is designed for fire departments, control centers, disaster management organizations, forestry and environmental authorities, municipalities, operators of critical infrastructure and security research institutions.