2008-2014
Robotics and GPU research
Embodied learning with the iCub robot, GPU acceleration, Aquila, CUDA teaching infrastructure, and ESA rover simulation.
Robotics folderWork
Four chapters: robots, cameras, coordinates, land.
Robots made perception answer to action. Cameras made it answer to real rooms. Coordinates made observations share a world. Tao made the same habit answer to land.
2014 onward
AI cameras and applied vision
Camera-as-computer prototypes, edge AI, synthetic worlds, and the bridge toward spatial perception.
AI cameras record
Spatial intelligence
Spatial intelligence
Helping observations share a physical frame: position, timing, topology, uncertainty, and physical context instead of isolated pixels.
Spatial record
Place
Tao
A place built slowly from rough land, where paths, water, structures, gardens, weather, and return become part of long-term care.
Tao recordHow it connects
Robotics tied perception to action; camera work moved it into real settings.
The chapters changed tools, scales, and methods: rover simulation, iCub learning, CUDA, edge cameras, shared coordinates, and finally Tao.
Robotics: perception tied to action and timing.
GPU: larger experiments became practical.
Edge vision: cameras met real settings.
Tao: land, water, weather, paths, and care at place scale.