Of the billions of operational Internet of Things (IoT) devices in the world at present, a large percentage are serving as a part of a widespread network of sensors. These sensor networks make things like precision agriculture, large-scale environmental monitoring, and smart cities possible. But these capabilities do not come without some downsides. In particular, keeping all of these individual units powered up and processing all of the data they produce is incredibly labor-intensive and very costly.
Typically, data is sent to cloud computing clusters for processing, which takes some of the load off of the devices, and renewable energy sources, like solar and wind, are often utilized to help keep batteries topped off. But excessive wireless data transmissions can drain batteries faster than renewable sources can keep up with. Furthermore, renewables are only intermittently available, and batteries still need to be replaced from time to time.
If you want something done right, do it yourself
In an ideal world, the IoT devices themselves would process the data they produce, and they would be completely self-reliant with respect to power. These goals are at odds with one another, however, so devices that meet them are few and far between. But a team led by researchers at Northeastern University has come up with a novel approach that could give more IoT devices these superpowers in the future. They have developed smart sensors that can process the data they produce onboard, and they can be powered wirelessly by radio waves.
The key to this work is a concept borrowed from condensed matter physics known as the Ising model. The team leveraged this principle to build what they call SPIN (Sensing Parametric Ising Nodes). SPIN sensors use the dynamic behavior of coupled oscillators to interpret multiple simultaneous inputs — similar to how the human brain processes complex information — allowing them to make more accurate and informed decisions in real time.
The future is looking smarter
This approach offers a number of potential benefits. First, because these sensors do not rely on lithium batteries, they sidestep the environmental issues and maintenance burdens associated with battery-powered devices. Second, their ability to locally analyze and react to captured sensor readings, like temperature spikes or the presence of hazardous chemicals, means they can reduce unnecessary data transmissions, thereby conserving energy and increasing reliability.
A working prototype has already demonstrated accurate temperature threshold sensing using this new method. Future iterations are expected to handle additional parameters such as humidity, light, and even the structural integrity of buildings. The researchers also envision their use in cold-chain logistics, power plants, and other areas where reducing emissions and waste is a high priority. With tens of billions of IoT devices in operation, this work could make a big impact as we build a more intelligent future.