The use of radiation in modern society is increasing, especially in medical radiotherapy and nuclear energy. Identification and characterization of radiation fields thus become crucial for ensuring safety. Current measuring instruments are often tailored to one type of radiation, which complicates the radiation characterization. Pixel detectors are suitable for imaging various radioactive sources, but they lack software tools to recognize the composition of the radiation field or to map its directionality.
The project aims to develop a tool for simulating events detected by a pixel detector in a mixed radiation field. This tool is essential for streamlining the process of characterizing potentially hazardous radiation sources in the environment through artificial intelligence methods.
For the simulations we are using Allpix2 software developed in CERN, and a machine learning model to improve the simulations.
detector, in comparison to an Allpix2 simulation