The Signal
UKRI backs Bonato project for real-time quantum-sensor experiments
UKRI selected Heriot-Watt professor Cristian Bonato for a three-year fellowship project to develop AI that changes which quantum-sensor measurements it takes as an experiment runs. The project is part of a 20-fellow funding round backed by more than £39 million.

Funding for an adaptive experimenter
AI is set to choose which measurements a quantum-sensor experiment takes next, rather than leaving researchers to review results after a run and then plan another round. UK Research and Innovation (UKRI) named Heriot-Watt University professor Cristian Bonato as lead of the project in its announcement on 16 November.
The three-year project, called “Developing a real-time AI-enhanced experimenter for quantum sensing,” is one of 17 Turing AI Pioneer Interdisciplinary Fellowships. UKRI’s funding round covers 20 fellows and is backed by more than £39 million in government funding.
Decisions during an experiment
The project will develop an AI system that selects measurements, learns from their results and adjusts its next steps while an experiment is under way, according to Heriot-Watt. The aim is to concentrate on measurements that provide more information and skip less useful ones; the university says that approach could speed up quantum-sensor experiments.
Researchers typically analyse measurements after an experiment ends, decide what to test next and then collect further data, the university said. The proposed system instead would work directly with a sensor and adapt the experiment as information arrives.
Addressing ambiguous results
The system is also intended to handle cases where the same results could fit more than one scientific explanation. In those cases, it could choose another measurement to help distinguish between the explanations, Heriot-Watt said.
Quantum sensors use individual particles, including electrons, to detect very weak signals; the university gives signals from single molecules as one example. It says these sensors can probe information conventional technologies may struggle to measure, although experiments can be slow and results challenging to interpret.
Control and potential uses
The work will also examine how AI might help researchers control individual electrons and atomic nuclei more precisely. Heriot-Watt says that improved control could contribute to quantum memories, which store quantum information using single particles and are building blocks for future secure quantum communication networks.
The university identifies healthcare and quantum communications as potential application areas. Improved sensors could help investigate biological processes at difficult-to-access scales, while more precise identification and control of individual particles could support more reliable quantum memories, it said.
UKRI’s announcement says the Turing AI Pioneer Interdisciplinary Fellowships began in October 2026 and will run for three years. The wider cohort also includes three Turing AI Global Fellows, whose fellowships are due to start in 2027 and run for five years.
Sources
Related stories
- CMOS Titanium Nitride Becomes a Compact Quantum-Dot Sensor
- Jesús Fuentes Aguilar Proposes a Quantum Sensor That Resets After Gain
- Jin, Ye and Ma Propose Faster Quantum Noise Spectroscopy
- UK Sets 15-Year Route to Put Quantum Systems in Space
- Diamond spin spirals sustain collective dynamics, paper reports