Last week, I hosted four experts for an hour-long discussion on RF sensing, which I think will become an essential tool for adding more context to IoT products. Each speaker introduced one of four technologies, after which we delved into a conversation about those technologies’ respective limitations as well as privacy and the use of new spectrum for sensing in the 60 GHz band.
You can find the full video here. Below is a quick summary of the key points — and a glimpse of the future.
We started with an introduction to Wi-Fi 802.11bf, by Chris Beg, analog designer with Cognitive Systems. He’s part of the IEEE standards-setting effort for Wi-Fi sensing, which aims to turn Wi-Fi devices into sensors that work with other Wi-Fi devices to give a rough picture of how things are moving in a room. The technology takes in data about the disruption of Wi-Fi signals; it can measure movement and detail specific kinds of movement. How fine a resolution it can detect depends on what frequency the Wi-Fi signal is sent over. Beg told us that we’re not going to get a formal specification until 2024, but that the standard development effort is on schedule.
We started with Wi-Fi because an audience poll made clear it was the thing people were most eager to learn about. Ultra-wideband (UWB) was the runner-up, and for that we turned to Ramona Marfievici, IoT practice lead at Digital Catapult, who explained what UWB is and why we so often see it paired with Bluetooth. (UWB is probably most familiar to people as an ingredient in Apple’s AirTags.) Marfievici also broke down why you can’t use fixed UWB outside, and why it draws so much power. And she described how good it is at getting really fine-grained location information.
Marfievici was followed by Bill Flaherty, who is the head of hardware engineering at Very, an IoT development shop and sponsor of my show. Flaherty talked about how the cost of LIDAR sensors was coming down and what that means for devices that can “see” highly detailed topographical maps of an environment. It’s probably not going to be on your vacuum anytime soon, but it makes sense for some factory applications.
Finally, Marcus DaSilva, a distinguished engineer at NI, detailed radar, which included running through various different types of radar and which of them would be best suited to different use cases. He also shared what would be an exciting future for RF sensing, namely that we could use terahertz spectrum to detect specific molecules and in turn use terahertz frequencies and radios as a way to track specific ingredients in food or pollutants in the air. I probably won’t be having events about that for another five years or so, though.
But regardless of the type of RF sensing that gets deployed, it will only ever be as smart as the algorithms used to interpret the data. With that in mind, Flaherty said that if someone wants to actually build products that rely on this sort of sensing they need to hire a good data science team. Meanwhile, all of the panelists agreed that such data should be protected to ensure people’s privacy, and most agreed that the data would be processed locally and only the necessary insights would be sent to the cloud or shared with a company.
Together, we also laid out time frames for when some of this might hit the mainstream. As I’ve previously noted, a formal standard for Wi-Fi sensing will take until 2024, but some features have already been built into routers and other gear currently on the market. Apple already uses LIDAR in the iPhone, for example, and Flaherty thinks it will become more common in drones before it hits smart home products. And the use of radar for wellness data will only continue. The same goes for UWB.
The session was deeply nerdy, and so if any of these issues are of interest you, I encourage you to watch it. I know that as a consumer, I’d love to see my devices get more context, which is why I hope to soon see some of these capabilities in the real world — in ways that protect my privacy.