Introducing Optimization-based
Control to Forest Surveillance

FirePuma can construct an uncertainty map from sensor data and autonomously guide drones to inspect the forest.

objectives

FirePuma was built with 3 main objectives.

map

Detect and Classify Events

How to detect and eliminate false data crowdsourced from the population or from sensors?

map

Maintain a Risk Map

How to update the support and probabiliy distribution associated with the uncertainty of existing fire?

optimal path

Optimal Surveillance Trajectories

How to design optimal trajectories to cover the most risky area in fully autonomous operation?

Videos and Pictures

Check the results obtain during the project.

Model Predictive Control for Route Planning

The algorithm uses a prediction horizon of the future trajectory but also a history of past positions to evaluate the quality of the path using the information of the uncertainty risk map.

gs

Model Predictive Control for Heterogeneous Route Planning

The algorithm uses a prediction horizon of the future trajectory but also a history of past positions to evaluate the quality of the path using the information of the uncertainty risk map.

gs

Unknown Target Pursuit

The evolution of the probability distribution for an unknown target can be incorporated into a Model Predictive strategy to allow for an optimized search. This is particularly suited for unauthorized drones or lost people.

Decentralized Controller with Local Information

The first solution was purely decentralized with the drone only requiring local measurements to perform the decisions and without global knowledge of the risk map.

Drone and LoRa Sensor Network

Integrated solution of a LoRa sensor network transmitting events to a node transmitting to the drone that executes event-driven missions.

Optimized Route in the Field

Final experiment in the field with the typical surveillance being conducted.

gs

Team

FirePuma has been developed in Institute for Systems and Robotics (ISR) that is part of IST-ID.

ISR logo IST-ID logo
16

journal papers

9

conference papers

19

MSc thesis

2

book

Events

Dissemination activities where the FirePuma team participated.

90s banner

Ep. 1392 Daniel Silvestre – Projeto FIREPUMA desenvolve rotas para vigilância e prevenção de fogos florestais através de deteção remota

Featuring in Antena 1 september 2022

Promotion activity to reach the general society regarding FirePuma research.

expo fct banner

FirePuma was present at ExpoFCT

Featuring in FCT-UNL April 2022, 2023, 2024

Biggest showroom of Science, Technology, Engineering and Mathmatics (STEM) activities for highschoolers in Portugal. FirePuma researchers were presented in all editions to reach the new generation of researchers with the scientific developments.

Final workshop banner

Final Workshop to Present Project Results

Featuring in DEEC @FCT-UNL january 2025

The final workshop was aimed at presenting the research developed within FirePuma.