Activity: Talk or presentation / Lecture › Invited talk at a scientific conference or institution
Description
Automating industrial transportation and production workflows offers significant potential for increasing productivity, lowering costs, enhancing safety, and delivering wider social and economic benefits, such as improving working conditions, addressing labor shortages, and driving GDP growth. From a technological perspective, however, realizing this vision still involves considerable challenges. It requires robust AI-driven perception, planning, and control algorithms that enable machines to operate reliably in weakly- or even uncontrolled environments. Furthermore, these machines must handle a wide range of tasks in complex 3D spaces: navigating and manipulating objects, often in collaboration with humans and other machines while maintaining continuous, 24/7 operation. In this talk, I will present our research and implementation results on navigation and pose-aware 3D object detection, along with real-world examples of successfully automated large industrial machines deployed in mining, forestry, and logistics.