Fredrik Sandin

Research

We work in both application-driven research projects where machine learning, artificial intelligence and computational physics are used to solve challenging problems, as well as in related fundamental research projects.

I'm interested in brain-inspired computing principles and neuromorphic technologies that offer fundamentally new ways to process information and implement learning for artificial intelligence (AI) purposes, for example in application domains requiring high energy efficiency such as the Internet of Things (IoT). The present computing systems and deep learning approaches are resource intensive and do not offer a sustainable path for large-scale AI in the digitization of society. 

More info: Google scholar profile, Semantic scholar profile.

Supervision

PhD students

• Mattias Nilsson, pattern recognition with DYNAP neuromorphic system. Supported by The Kempe Foundations
• Karl Ekström, natural language processing for fault severity estimation, KnowIT FAST.
• Rickard Brännvall (RISE), machine learning for automation of edge datacenters, AutoDC.
• Co-supervision of Gustav Grund Pihlgren, Saleha Javed, Carl Borngrund, Lars-Johan Sandström, and Muhammad Ahmer (SKF).

Graduated PhD students

• Jacob Nilsson, thesis: Machine Learning Concepts for Service Data Interoperability (2022).
• Kim Albertsson, thesis: Machine Learning in High-Energy Physics: Displaced Event Detection and Developments in ROOT/TMVA (2021).
• Siddharth Dadhich, thesis: Automation of Wheel-Loaders (2018, co-supervisor).
• Sergio Martin del Campo Barraza, thesis: Unsupervised feature learning applied to condition monitoring (2017).
• Blerim Emruli, thesis: Ubiquitous Cognitive Computing: A Vector Symbolic Approach (2014).

Postdocs

• Daniel Strömbergsson, Co-design optimization of neuromorphic condition monitoring sensor system.
• Saad Arif, starting in January.

Past Postdocs

• Sergio Martin del Campo Barraza (now adjunct senior lecturer in our group), unsupervised machine learning for automation of wind turbine condition monitoring system, two years funded by SKF.

Teaching

Examiner of

• Programming for Machine Learning (D0036E).
• Neural Networks and Learning Machines (D7046E).
• Neuromorphic Computing, upcoming course in the Applied AI program and Applied AI master.
• Programming for Scientific Computing, upcoming course in the Engineerng Physics and Electrical Engineering program.
• Master thesis courses in Engineerng Physics and Electrical Engineering (X7010E, X7011E, X0003E).

Since 2014, I am program director of the "Teknisk fysik och elektroteknik" program at LTU (Civilingenjör, 300Hp), co-directed with Andreas Almqvist.

Lectures also in D0032 Introduction to AI and D0028E Programming and Digitalization. Was examiner for E0003E Electric circuit theory, D0011E Digital design, D0017E Introduction to programming for engineers. Before 2010 I taught courses in physics, including lectures about mechanics, thermodynamics, waves, optics, quantum physics, and also some 3rd-cycle lectures in astrophysics, cosmology and quantum field theory at finite temperature (J. Kapusta's book).

Background

MSc diploma work in ATLAS at CERN (2001). PhD in Physics (2007, LTU, Swedish Graduate School of Space Technology). I received the 2004 “New-Talents” award for original work in theoretical physics at the International School of Subnuclear Physics in Erice. First postdoc at IFPA in Belgium (2008-2009), focusing on computational physics. A growing interest for the physics of brains and neuromorphic technologies made me shift research focus. I did a second postdoc in brain-like machine learning (2010-2011) at EISLAB, where I became assistant professor (until April 2016), associate professor (until March 2021) and presently work as professor in Machine Learning. I was awarded the Gunnar Öquist Fellowship from the Kempe Foundations in 2014 which included 3 MSEK for research. I served as a technical committee member of the SKF–LTU University Technology Center, as a research theme leader of the Intelligent Industrial Processes area of excellence, and as a member of the Applied AI Innovation Hub. I served in several faculty workgroups focusing on improving education processes. I am a member of ELLIS since June 2022.

Other

Some code and tools. Contact me if you are interested in code used in research articles that are not listed below.

• 3FCS code, link .This code was developed and used for the quark matter calculations in the papers about neutron stars with quark matter cores and related phase diagrams, including the highly cited paper in Phys. Rev. D.
• Femtolensing tool, link . Calculates the gravitational lensing signatures of low-mass (10 ¹⁴ -10 ¹⁷ kg) compact interstellar objects. This code was developed when working on the preon star hypothesis, see highlights below.
• N-dimensional random projection code, link . Result of early work on representation learning for cognitive computing.
• CBVS code, link. Early work on cognitive computing, related publications here and here.
• Dircheck, link . A useful tool for verification of file archives.
• Python code for distributed computing over ssh, link .
• ASOUND Matlab plugin, link .
• Online collaborative writing in Latex, link . Developed before the time of Sharelatex / Overleaf etc.

Some highlights

• ELLIS member.
• Gunnar Öquist Fellow award and grant from The Kempe Foundations, link .
• Dark matter neutron-star core hypothesis. Paper, paper, impact.
• ISSP award for an original work in theoretical physics, link (more about Prof. 't Hooft and Prof. Zichichi ).
• Preon star hypothesis. Paper , paper , observational predictions , Nature news , Phys. Rev. Focus , New Scientist .