The wireless sensor can think on its own without batteries or the internet, helping devices work and save energy in buildings and more.
Your home’s thermostat may seem basic, but it relies on a wireless sensor to decide when to activate heating or cooling—an example of how sensors power the Internet of Things (IoT). These devices allow systems to collect and share data, yet most still depend on lithium batteries and struggle with decision-making, such as identifying multiple threats before responding. Researchers at Northeastern University have used principles from condensed matter physics to embed logic directly into a wireless sensor tag—a development that could change the future of IoT and artificial intelligence.
Most wireless sensors depend on energy harvested from nearby radio waves or light, which can be inconsistent. They typically cannot process the signals they detect or perform computations before transmitting data to a reader.
The passive wireless sensor tag developed in this work can perform computations on multiple environmental parameters. This enables decision-making within wireless sensor networks and reduces dependence on cloud resources.
Using the Ising model—a concept from physics adapted for quantum computing—researchers created a passive wireless sensor capable of decision-making similar to the human brain. Called SPIN (Sensing Parametric Ising Node), the component can respond to multiple data sources and make decisions.
SPIN can sense and perform functions not possible with other passive wireless sensors. This could help reduce gas emissions and energy use in buildings and power systems, and reduce waste in cold-chain systems.
With billions of sensors expected to support internet-connected devices by the end of 2025, embedding logic into passive wireless sensors will allow AI and machine learning systems to process data locally and reduce the need for cloud resources.
A prototype can detect changes in temperature. Future versions are expected to measure parameters such as humidity, light, and structural soundness in buildings and bridges. These sensors may also detect human presence or identify environmental patterns, including chemicals.
Each passive wireless sensor can compute and make decisions based on local inputs, helping central systems gain better insight with less dependence on cloud computing.
Reference: Nicolas Casilli et al, Programmable threshold sensing in wireless devices using Ising dynamics, Nature Electronics (2025). DOI: 10.1038/s41928-025-01392-4
