Max Robotics — Global Service Robot Directory + US Certification

About

Martin Gerdzhev’s research bridges the cutting edge of autonomous perception with the life-saving practicality of urban search and rescue (USAR). His most impactful work centers on **3D semantic segmentation for autonomous systems**, where he developed TORNADO-Net, a neural network that fuses bird’s-eye and range-view projections to parse LiDAR point clouds with remarkable accuracy. This multi-view total variation approach, enhanced by a Diamond Inception module, has earned over 89 combined citations, establishing Gerdzhev as a key contributor to scene understanding for robotics and self-driving vehicles. Equally compelling is his pioneering work in **canine-assisted robotics**, where he co-designed DEX (Drop and EXplore), a small marsupial robot deployed by trained search dogs to explore rubble inaccessible to larger machines. This concept, detailed in his 2010 paper (21 citations), reimagines human-robot-animal collaboration, enabling faster, safer survivor detection. Gerdzhev’s career uniquely spans high-performance deep learning architectures and field-tested rescue systems, demonstrating a rare ability to drive both algorithmic innovation and tangible humanitarian impact.