Laboratory tools move measurement closer to the moment it is made
An X-ray data network and a battery-free folding implant follow the same design direction in very different settings: reducing the wait between measurement and usable information.
Science··Midday
Without waiting for the beamline's aftermath
According to Argonne National Laboratory's announcement, a physics-aware neural network called DONUT can interpret X-ray nanodiffraction data while an experiment is still running. This kind of imaging follows what is happening inside materials through many measurements from a small area; on the conventional route, making sense of the data can take weeks or months. The report says DONUT needs no labelled training examples and returns results hundreds of times faster than the established analysis route. That speed claim is the team's comparison with the conventional method. The change includes faster computation and lets a researcher see what they are looking at before the sample leaves the beamline. The system can learn from data collected at the start of a run and be adjusted to predict different quantities. In practical terms, a late explanation can give way to a change of direction during the experiment.[1]
Moving information under the skin without a battery
On the medical-device side, a folding tool called MiFi deals with a different limit of time and power. According to Medical Xpress, the device from Imperial College London and the University of Southern California begins as a 0.3-millimetre sheet and unfolds to 2.1 by 2.1 centimetres; it uses near-field communication for power and readout. The report says it can transmit measurements including heart and breathing rates, temperature, tissue acidity and lithium levels. The point is the possibility of communication for a device under the skin without the volume and replacement needs of a battery. The situation described by the report remains at the laboratory stage: the device was implanted in anaesthetised rats and tested with pig skin and tissue samples, and no person has received one. MiFi is therefore not a ready clinical tool. It is a prototype showing how measurement and power transfer might be combined in a very small surface.[2]
Speed does not replace judgement
DONUT and MiFi are neither the same object nor the same use case. One aims to make materials data understandable while a large beamline experiment continues; the other connects a small measuring surface under the skin to an external reader. Their shared direction is to make data interpretable closer to the moment it is produced. The two tools have distinct use contexts and stages of readiness. Argonne's performance comparison is the team's comparison with its established route, while the report is explicit that MiFi needs further work before it can be used in people. Those limits do not diminish the tools' value. They distinguish their stages: one tries to return feedback to an experiment's flow, the other to combine power and measurement in a small implant. The common question today is when information, however quickly it arrives, is ready for use.[1], [2]
