Machine learning of detector signals for nuclear experiments

Engineering Programme / Master in Physics – Degree Project (Exjobb) 30 ECTS

In nuclear experiments we collect detector signals from the interaction of subatomic particles with matter. The precise knowledge about the signal characteristics is necessary to deduce the important observables from these nuclear interactions, such as particle energy and time of arrival.

In this project the student will obtain a large set of acquired signals from different particles ranging from alpha to heavy fission fragments. These signals were obtained at a recent experiment done at an intense neutron field in the Institut Laue-Langevin (ILL) nuclear reactor site in France.

The main project goal is to create a framework and analyse the signals to improve the extraction of nuclear observables. More specifically the tasks are:

  • Study the variations of time pick off as a function of signal height and rise time: The time of particle arrival is shifted as a function of amplitude and rise time (so called walk effects). This highly depends on the method of time extraction. We would like to study the different times pick off methods and their respective delays.
  • Simulation of various signals: This is needed to properly disentangle the so-called walk effect from any physical time differences. By artificially constructing signals one can understand the optimal setting for each algorithm.
  • Implementation of machine learning algorithms: to learn how the detector signals vary and how one can learn about the rise-time and amplitude changes, especially if they can be correlated with systematic time fluctuations.

This is a unique opportunity to analyse nuclear experiments, improve the detection capabilities. Programming skills are required, along with a curious mind.

The codes for analysing the signals exist in the ROOT C++ framework. However, one can also choose another programming language if so pleased.

Start date: January 2022

Ali Al-Adili (
Division of Applied Nuclear Physics – Department of Physics and Astronomy, Uppsala University

Senast uppdaterad: 2021-12-20