A Search and Detection Autonomous Drone System: From Design to Implementation

Utilizing autonomous drones or unmanned aerial vehicles (UAVs) has shown great advantages over preceding methods in support of urgent scenarios such as search and rescue (SAR) and wildfire detection. In these operations, search efficiency in terms of the amount of time spent to find the target is crucial since with time the survivability of the missing person decreases or wildfire management becomes more difficult with disastrous consequences. In this work, we consider the scenario where a drone is intended to search and detect a missing person (e.g., a hiker or a mountaineer) or a potential fire spot in a given area. To obtain the shortest path to the target, a general framework is provided to model the problem of target detection when the target’s location is probabilistically known. To this end, two algorithms are proposed: Path planning and target detection. The path planning algorithm is based on Bayesian inference and the target detection is accomplished by using a residual neural network (ResNet) trained on the image dataset captured by the drone as well as existing pictures and datasets on the web. Through simulation and experiment, the proposed path planning algorithm is compared with two benchmark algorithms. It is shown that the proposed algorithm significantly decreases the average time of the mission.
Note to Practitioners
—This article is motivated by the need for an efficient path-planning algorithm for drones during specific SAR operations. In particular, situations where someone is lost in a snow-covered hike and a fire spot that is in its initial levels are of interest. In fact, since the target location is not known, it is required that the UAV be able to efficiently search the entire area until it finds the target in the shortest possible time. The proposed Bayesian framework along with the ResNet learning algorithm shows an efficient performance in terms of average time duration and accuracy, respectively. The framework developed in this paper can be extended to a multi-UAV scenario where UAVs coordinate to optimize the overall performance.