Heriot-Watt develops adaptive quantum sensing research

Heriot-Watt develops adaptive quantum sensing research

Heriot-Watt researchers will develop adaptive measurement controls for quantum sensors. Professor Cristian Bonato’s three-year project will investigate artificial intelligence that selects quantum measurements using experimental feedback, with the aim of improving measurement efficiency and precision.


IN Brief:

  • Professor Cristian Bonato is receiving a three-year Turing AI Pioneer Interdisciplinary Fellowship.
  • The project will investigate AI systems that select measurements in response to experimental results.
  • The research targets more efficient quantum sensing, although no commercial sensor performance improvement has yet been established.

Heriot-Watt University will investigate how artificial intelligence can control quantum sensing experiments, selecting measurements in response to experimental results rather than relying exclusively on predetermined observation sequences.

Professor Cristian Bonato, a specialist in spin-based quantum technologies, has received a three-year Turing AI Pioneer Interdisciplinary Fellowship to develop an adaptive experimental system intended to improve the efficiency and precision of quantum measurements.

The project forms part of a UK Research and Innovation programme supporting 20 Turing AI fellows through two complementary schemes, with more than £39 million in combined funding. Bonato is among 17 researchers selected for the interdisciplinary fellowship programme.

The funding total covers the wider fellowship awards and is not the amount allocated specifically to Heriot-Watt. The university has not disclosed an individual award value in its announcement.

Quantum sensors exploit measurable properties of quantum systems to investigate physical quantities such as magnetic fields, time, and changes within materials. Their performance depends on the interaction between the sensing element, the surrounding environment, and the techniques used to extract information from the system.

Measurements are not always equally informative. The experimental conditions chosen for one observation may provide useful evidence about a particular property while contributing relatively little to understanding another.

Conventional procedures can involve repeating a sequence of measurements under predetermined conditions. Researchers then compare the resulting data with mathematical models to estimate the properties of the system under investigation.

This approach provides a structured basis for collecting observations, but it can also require measurements that become less valuable as information accumulates.

Bonato’s research will investigate whether artificial intelligence can improve that process by deciding which measurement to perform next.

An adaptive system could compare different explanations of the observed behaviour, identify areas of uncertainty, and select experimental settings that help distinguish between competing possibilities.

Each new observation would then be used to update the controller’s understanding of the system. The subsequent measurement could be selected according to the information already obtained rather than following a fixed sequence.

The underlying concept is related to adaptive experimental design and statistical estimation. In these approaches, observations are evaluated according to the information they provide, allowing subsequent measurements to concentrate on unresolved questions.

Quantum sensing introduces additional constraints because the system being measured may change during observation. Environmental disturbances, finite coherence times, and measurement noise can limit the information available from individual experiments.

The control system must therefore account for uncertainty rather than simply adjusting its measurements according to the latest observed value.

A noisy observation could otherwise lead the controller towards an unsuitable experimental setting, reducing measurement efficiency or producing an incorrect interpretation.

Bonato’s project will examine how artificial intelligence can make these decisions during the experiment itself. The intended system would connect measurements, interpretation, and the selection of subsequent operations in a continuous feedback process.

Such feedback introduces computational requirements. The controller must process information quickly enough to influence the experiment while maintaining an appropriate representation of uncertainty.

Its effectiveness will depend on the measurement hardware, the complexity of the quantum system, and the processing time required to select each new experimental condition.

The university expects the research to contribute to more precise quantum sensing, with potential applications including healthcare and secure communications.

Those applications remain prospective. The fellowship has not produced a newly qualified sensing device or a demonstrated improvement across a defined range of commercial instruments.

Heriot-Watt already undertakes research in quantum photonics and spin-based sensing. Its work includes experimental and theoretical approaches to quantum measurement, providing a foundation for investigating more sophisticated control methods.

The university is also involved in four of the UK’s five national quantum technology hubs, supporting research into different aspects of quantum technology.

Moving an adaptive measurement method from laboratory research into equipment would require additional development. The software would need to operate reliably with specific sensing hardware, while measurement results would need to remain repeatable under changing environmental conditions.

Calibration and traceability would also be important. If the controller changes the measurement procedure during operation, engineers must still establish how the resulting measurements can be compared, validated, and interpreted.

A faster sequence of observations is not necessarily more accurate if the control process introduces bias or fails to account for measurement uncertainty.

The research will therefore need to assess both the amount of information obtained and the reliability of the resulting estimates.

Bonato’s fellowship began within the October 2026 programme and will run for three years. Its initial objective is to establish an experimental approach in which artificial intelligence improves measurement selection through feedback.

The resulting methods could inform future sensing equipment, but their practical value will depend on demonstrated improvements in measurement precision, efficiency, and reliability under defined operating conditions.


Stories for you


  • Samsung expands NVIDIA partnership into MLCCs

    Samsung expands NVIDIA partnership into MLCCs

    Samsung Electro-Mechanics has extended its semiconductor partnership with NVIDIA further. The companies have amended their substrate supply agreement and broadened cooperation involving high-reliability multilayer ceramic capacitors for accelerated computing platforms.


  • Vishay extends high-voltage capacitor range

    Vishay extends high-voltage capacitor range

    Vishay has expanded its high-voltage aluminium electrolytic capacitor product range. New 550V and 600V versions of the 259 PHM-SI series provide additional options for DC bus circuits in industrial drives, energy systems, and high-power conversion equipment.