About this role
Internship | Teaching a Drone Where to Look for Better 3D Reconstructions TNO Keywords: 3D-vision, AI, computer vision, machine learning , Python Contract Type: Type: Internship Location: Den Haag Education Level: Education: Master (EQF 7) Published on: Published: 23/09/2026 Status: Open Apply before: 01/10/2026 Hours p/wk: 32 Make your mark on our time. Become an intern at TNO! About this position At TNO, the Netherlands Organisation for Applied Scientific Research, researchers work on turning cutting-edge science into technology that matters in the real world - from healthcare and energy to defense and mobility. Within TNO's work on autonomous systems, one recurring challenge is teaching machines to see, understand, and act in complex, unpredictable environments - the same kind of everyday messiness that makes tasks trivial for humans but remarkably hard for robots. This internship sits within that effort, focusing on drones that can autonomously make a good 3D reconstruction. You will study a drone that plans its own flight path for 3D scanning, deciding in real time where to fly next based on what it has already captured, rather than following a fixed grid or orbit and hoping the result reconstructs well. Using fast, lightweight 3D foundation models (like VGGT or MapAnything) that estimate rough scene geometry from the live camera feed, the drone builds a running sense of which parts of the scene are well covered and which still need better angles, distances, or overlap - without ever needing to build the slow, high-quality 3D Gaussian Splat reconstruction mid-flight. On top of this general coverage strategy, the drone can also be told, in plain language, about specific objects of interest ("the red fire hydrant"), which it locates using open-vocabulary detection models and then prioritizes for extra close-up passes until it has captured enough varied viewpoints to reconstruct that object wel...