@article{article_1680411, title={Early Fire Detection Mobile Robotic System With Hybrid Locomotion}, journal={Black Sea Journal of Engineering and Science}, volume={8}, pages={1111–1120}, year={2025}, DOI={10.34248/bsengineering.1680411}, author={Sucuoğlu, Hilmi Saygın and Böğrekci, İsmail}, keywords={Early fire detection, Faster R-CNN machine learning model, Fire detection robot, Hybrid locomotion, Mobile robotic system, Path planning and obstacle avoidance}, abstract={In this study, a mobile robotic structure with fire detection and hybrid locomotion capabilities designed and developed. The hybrid locomotion system is an adaptive three-wheeled structure, and it has been structured to provide obstacle climbing and linear motion. The paper puts forward a structure for obstacle avoidance and path planning named "Direction Based Angle Computation". The system is designed to categorize obstacles as either negligible or surmountable, with this classification determined by the height and shape of the obstacles. The objective of the "Fire Search and Find" and "Fire Detection" systems is to identify potential fire locations and calculate the associated probabilities. Experimental tests are conducted for the mechanical structure and architecture of robotic systems. The experimental test results demonstrated that the motion systems have proficiency in both rolling-climbing and linear motions. The Direction Based Angle Computation approach is a proper methodology for the tasks path planning and obstacle avoidance. The proposed fire detection algorithm with the usage of Faster R-CNN machine learning model, has been shown to determine the probability of a fire source with 93% accuracy.}, number={4}, publisher={Karyay Karadeniz Yayımcılık Ve Organizasyon Ticaret Limited Şirketi}