Somalytics turns paper into cheap proximity sensors

This story was originally published in my Sept. 30 newsletter. 

This week’s newsletter is all about sensors.

When I evaluate technology, I often look at whether or not it changes the cost, power consumption, or performance of a typical process by a significant factor. One of the easiest ways to do that is through new materials or better silicon. And sometimes new materials lead to better silicon.

Somalytics has used new materials — specifically carbon nanotubes — to create a new type of sensor manufactured in a way that enables them to be incredibly cheap and incredibly power efficient, and to sense movement and items more efficiently than any existing counterparts.

The Seattle-area company was created in November 2021 based on research from the University of Washington. It recently raised $1.9 million to help it bring its process into commercial production.

Somalytics’ sensor can detect proximity and elements such as solids or liquids based on the disruptions those elements have on nearby electric fields. The company has figured out a way to coat paper in a thin layer of carbon nanotubes (a newish material!) and tear it in such a way that it can detect very minute disruptions in the electrical field around the sensor. When a nearby object disrupts the electric field, the object closes the circuit and generates a signal that the sensor can detect. It’s similar in principle to both haptic sensing, where placing your finger on a pad closes a circuit, and measuring radio frequency disruptions to detect movement.

With haptic sensors, the size of the field created to measure disruption is limited (mostly because larger sizes are expensive to make). That means the further away the item you want to detect, the larger and more expensive the sensor needs to be. But Somalytics’ sensor can detect movement farther away than a haptic sensor, and in 360 degrees. The tearing technique pioneered by Somalytics creates hundreds of micro tears in the paper, which creates a surface area large enough to be able to measure disruption. To visualize the difference in surface area, imagine the difference between the perimeter of a smooth coastline and one riddled with fjords, where the length of the coastline corresponds to how far away it can sense a disruption.

The result is a sensor with a large surface area that, since it’s made using a roll-to-roll paper manufacturing process as opposed to traditional semiconductor manufacturing, is relatively cheap to produce. It also consumes a mere 135 milliwatts per module.

The Somalytics sensor can replace Passive Infrared (PIR) sensors for detecting movement as well as the ultrasonic sensors used in industrial settings to measure the levels of gases, liquids, or even solids. Barbara Barclay, the CEO of Somalytics, told me the company is working with customers that make products for bathroom facilities to replace the PIR sensors that sit beneath faucets to turn the water on in response to hand movement, or in paper towel dispensers that dispense towels in response to a hand wave.

PIR sensors, after all, can get dirty and thus become less effective over time. And they don’t work very well with people who have dark skin tones. The Somalytics sensor doesn’t have either of these issues. Moreover, it can detect motion up to 200 millimeters (almost 8 inches) away, meaning someone doesn’t have to get super close to the sensor to trigger it. Obviously this could become a problem (I’m imagining a paper towel dispenser shooting out said towels every time someone walks too close to it).

In the meantime, industrial customers in consumer packaged goods and the automotive industry are interested in using the Somalytics sensor to detect how much liquid is in a container, which would replace existing ultrasonic sensors or other mechanisms used to measure capacity. They could be used in food settings to ensure the right amount of liquids are dispensed into a package or in automobiles for detecting how much wiper fluid or even gas is left. Because these sensors leverage paper, they are cheap to make and can be used more widely than more expensive options.

Barclay didn’t tell me exactly how much the Somalytics sensor costs, but did say the company is manufacturing 10,000 sensors a month on a machine that’s about six feet by six feet. The goal is to be making 10 million sensors a year by the end of 2023, at which point the company will seek additional funding.

While all of the aforementioned use cases are ones where the Somalytics sensor could change the economics and performance of motion sensing enough to drive adoption, it could also provide a completely new use case and measurement. The same way the sensor can detect water filling a glass means it can detect minute changes in water in other areas — including the human eye.

Eyeballs are filled with a water-based goo (called vitreous gel) and the Somalytics  sensor can measure the status of that gel accurately enough to detect eye movements. This could be of tremendous interest to companies in the health tracking space, which might want to understand sleep patterns, or in the automative sector, where gaze detection is an important indication of driver attention.

Barclay didn’t get into all of the specifics, but she said the company is already working with customers to put the sensor in products. For now, Somalytics has designed a package that includes the sensor and a microcontroller with the algorithms necessary to make it easy for customers to test the capabilities of the sensor before designing it into their own products.

And if it catches on, we could see a world of interfaces that allow users to merely point at screens to, say, order food or catch an elevator. They could even be used in light switches to turn on lights, or in eyeglasses to detect gaze and then provide more information. Better sensors mean better devices, and maybe even new interfaces.