Dandelions have evolved to disperse their seeds more than a kilometer in the air.

Researchers from the University of Washington want to give sensors that kind of distance, in a way that supports agricultural and environmental-monitoring applications.

Usually it's not the best idea to drop valuable wireless sensors from great heights. Led by Professors Shyam Gollakota and Vikram Ayer, the team at UW, however, did just that, creating a tiny sensor-carrying device that can be blown by the wind as it tumbles toward the ground.

Like dandelion seeds, the sensors float in the breeze. The device, about 30 times as heavy as a 1-milligram dandelion seed, can travel up to 100 meters on a windy day.

To keep the devices light and to ensure that the sensors landed with the solar panels facing skyward, the UW engineers needed to mimic the dandelion's shape.

“The way dandelion seed structures work is that they have a central point and these little bristles sticking out to slow down their fall. We took a 2D projection of that to create the base design for our structures,” said lead author Vikram Iyer, a UW assistant professor in the Allen School . “As we added weight, our bristles started to bend inwards. We added a ring structure to make it more stiff and take up more area to help slow it down.”

With laser micromachining, Iyer and the team could test out a variety of patterns and sizes.

The researchers tested 75 designs, some of which are shown here in yellow. (Photo: Mark Stone/University of Washington)

Sensor data like temperature, humidity, pressure, and light can be shared from a distance of 60 meters. The engineers designed the lightweight, flexible circuits and electronics to include a capacitor, a device that stores some charge overnight.

During one test, a drone dropped sensors from a height of 20 meters and sent the sensors about 100 meters across a nearby parking lot. (See the video below).

“This is just the first step,” Iyer said. “There are so many other directions we can take now — such as developing larger-scale deployments, creating devices that can change shape as they fall, or even adding some more mobility so that the devices can move around once they are on the ground to get closer to an area we’re curious about.”

In a short Q&A with Tech Briefs below, Iyer talks more about the advantage and disadvantages of sending sensors out, seed-style.

Tech Briefs: What inspired the choice to imitate the dandelion? (The idea of, kind of, scatter-shooting sensors feels slightly counter-intuitive to me!)

Prof. Vikram Ayer: If we think about a dandelion seed from an engineering perspective, it has some pretty amazing capabilities. These little plants can't even move, but they've evolved to be able to let their seeds disperse for up to a kilometer in the right conditions. This is exactly what we'd like to do for automating deployment of wireless sensor networks. If we want to take sensor measurements over a really large geographic area to do environmental monitoring for agriculture or climate change studies, this can be very time consuming and expensive, or even dangerous in some remote locations. In this work we instead look to dandelion seeds for inspiration to automate this process by creating sensors that can disperse in the wind.

Tech Briefs: How much can the sensor’s dispersal path be controlled? What are the advantages and disadvantages of sending out the sensor “Seeds” somewhat randomly?

Prof. Vikram Ayer: To get good coverage over an area, we actually look to nature again. Plants can’t guarantee that where they grew up this year is going to be good next year, and the natural variation between seeds allows some to travel farther away to hedge their bets. We take the same approach and design a whole array of different structures that float in the air for different periods of time. This means that even in the same wind conditions we can make sure some of them land closer and others travel farther to get uniform coverage over an area. As a next step, we're exploring ways we can change the shape of these structures mid-flight to get even finer grained control.

Tech Briefs: Wouldn’t it be important to know the location of each sensor? How could that be done?

Prof. Vikram Ayer: We can vary the design of the sensors to help get uniform coverage over an area. We've also shown multiple techniques for wireless localization before that we've used to track bumblebees , murder hornets , and small objects around the house or in a hospital  that we hope to integrate with this platform. We can do this by looking at things like the wireless signals strength and comparing the signals we get on multiple antennas to figure out the angle to the sensor and triangulate its position.

Tech Briefs: For backscatter, where would the transmitted signal come from? Would each sensor have to be interrogated separately?

Prof. Vikram Ayer: To transmit the signal and read back the data we build an access point with a radio transmitter and receiver, similar to a Wi-Fi router. One of the cool things about this work is that we show a single access point can communicate with one of our sensors up to 60m away, and we show this with experiments using access points on the ground communicating across a soccer field. We could also use a drone to carry this same setup to read the data, or use a hybrid setup with stations on the ground and drones depending on the deployment scenario. We explore multiple strategies to communicate with many sensors, for example by leveraging the fact that they're power harvesting and will start up at different times, and also adding time delays to make sure their transmissions don't interfere. We can also integrate our prior work on backscatter protocols  that can scale up to support many devices for improved performance in future versions

Tech Briefs: What applications is this capability most valuable for?

Prof. Vikram Ayer: This technology could be useful for all kinds of environmental monitoring applications where you want to spread sensors over a large area. For example for precision agriculture, environmental monitoring for climate change especially remote, hard-to-reach areas like forests and glaciers. Another important part of this work is we show how we can design these tiny, wireless computing and sensing devices with programmable, general purpose computing devices. This allows anyone with a computer science or engineering background to build on our system and customize the core computing and sensing platform for other applications like wearable sensors, medical implants, and microrobots.

Tech Briefs: What was it like to test this, and what was most memorable when you tested this?

Prof. Vikram Ayer: One of the really cool features of emulating the design of a dandelion seed is that it always falls with the same side facing up. Even if you drop it from upside down, you can see it flip over in mid-air to correct itself. This is actually very important for our design because it makes sure that our solar cells face upwards and can gather sunlight to power our battery-free sensor.

Tech Briefs: What’s next?

Prof. Vikram Ayer: In addition to the things mentioned above like designing ways to change the shape of the structure as it falls, wirelessly localize it, and explore things like biodegradable materials to make these devices more sustainable and prevent them from polluting the environment, this work is part of our broader vision of creating the Internet of bio-inspired and biological things. Specifically, there’s a pretty big gap between biological systems and the capabilities of current IoT and embedded systems, which are much larger and heavier and most can’t move around. Instead, imagine, if we can create tiny battery-free wireless devices that can move around, and in fact float in the air similar to dandelion seeds. If we could do that we could deploy hundreds of sensors in the wind in remote, hard-to-reach areas like forests, glaciers. Or if we create wireless sensors that are so small , then we can also start attaching them to tiny insects like bees, beetles and murder hornets  then we can use these sensors to study their behavior in the wild. If we can go a step further to integrate actuators with these wireless sensors, we can enable them to move around freely and build insect-scale robots .

What do you think? Share your questions and comments below.

Also: Read our "5 Ws" on the dandelion-inspired achievement.

Learn about 5 High-Tech Materials That Sense and Detect