AI & ML Practical Magic

An AI taught itself how to perform complex crystal alignment just by looking at pictures, without being taught a single law of physics.

April 14, 2026

Original Paper

Autonomous Diffractometry Enabled by Visual Reinforcement Learning

J. Oppliger, M. Stifter, A. Rüegg, I. Biało, L. Martinelli, P. G. Freeman, D. Prabhakaran, J. Zhao, Q. Wang, J. Chang

arXiv · 2604.11773

The Takeaway

This replaces years of expert human training with a machine that 'just gets' crystallography by recognizing abstract patterns in X-ray diffraction. It proves that neural networks can master hyper-specialized physical tasks without needing to understand the underlying textbook equations.

From the abstract

Automation underpins progress across scientific and industrial disciplines. Yet, automating tasks requiring interpretation of abstract visual information remain challenging. For example, crystal alignment strongly relies on humans with the ability to comprehend diffraction patterns. Here we introduce an autonomous system that aligns single crystals without access to crystallography and diffraction theory. Using a model-free reinforcement learning framework, an agent learns to identify and naviga