From incidental connections to near incomprehensible unknowns in the universe, Dr James Alvey is enjoying returning to Cambridge.
“I never met him; I did see him in Sainsbury’s occasionally,” says James, when asked about Professor Stephen Hawking. It is an inevitable question, as James is a Cosmologist at Gonville & Caius College and was previously an undergraduate student at the University of Cambridge.
Professor Hawking was a Fellow of Caius for 52 years until his death in 2018. James became a Caius Fellow and College Lecturer in Physics in October 2024.
The Senior Kavli Fellow in Gravitational Waves at Cambridge’s Institute of Astronomy read Mathematics at St John’s College from 2014 and a PhD in Theoretical Physics at King’s College London followed. His PhD looked at dark matter, cosmology, dwarf galaxies and the Milky Way and the universe.
James next went on to a post-doctoral research position at the Grappa Institute at the University of Amsterdam changing field somewhat towards data analysis and machine learning.
“For me the connection to Hawking’s work is very direct in the sense that I think a lot about black holes and how we can learn something about them,” James says.
“During my post-doc, I started looking a lot more at gravitational waves, mainly on the data analysis side. In other words, how do we optimally analyse the data we get in from detectors and get all the science that we want out of them? I am mainly interested in how we can achieve this using the huge developments in machine learning and AI which have come about just in time for the next generation of detectors.”
James’ main role is to interpret the data to learn about astrophysics and cosmology and ask questions such as: is that a gravitational wave? If it is, what sort system did it come from? He uses techniques in Bayesian inference to solve the inverse problem of going from the detector data to properties of the systems.
He says: “The amazing thing is that we can even make observations which connect directly to that incredibly violent event which happened somewhere in the universe. The main way to do this is via an interference pattern at an experiment such as LIGO (the Laser Interferometer Gravitational-Wave Observatory).
“That measurement is incredibly hard to make. The change in distance is unbelievably small – it’s like measuring the distance from here to the moon, to less than width of a human hair.
“It’s an incredibly highly precision science. It’s a huge technological development to make these gravitational wave systems the dominant thing you can see. We thought gravitational waves existed since the early 1900s, but it took about 100 years to build a good enough system to actually see them.”
James’ is also part of an international effort with the next generation of detectors, which are under construction. The LISA (Laser Interferometer Space Antenna) experiment is a European Space Agency project, scheduled for completion in 2035, making the work quite time sensitive.
“Space is obviously a big vacuum and if we put an experiment like this in space, we can open up totally different classes of objects that we might be able to see,” James says.
“This is a big challenge though – it’s in space! We need to have very good data analysis plans now, before it gets built.
“In the LISA context, my research mostly concerns how to use machine learning techniques to speed up the data analysis pipelines, but still maintain the scientific fidelity we want. Can machine learning accelerate that process and how? If it can, can we trust it? We want to do precision science with precision statistics. It’s really important if I put a machine learning algorithm in there, that we understand why it does what it does and that we can trust the answers.”
It is a long way from the aisles of a Sidney Street supermarket.
