Q&A: DecaWave’s Ciaran Connell on UWB, Indoor Location, Robots, and IoT
Knowing the position of a person or device can be quite useful in terms of the Internet of Things (IoT). However, determining that information indoors, where GPS is unavailable, can become a challenge. Unfortunately, GPS isn’t practical in many applications, such as very small, wearable devices. On top of that, many devices (e.g., robots) will only be used indoors, making GPS a bad choice for location information.
I talked with DecaWave’s Ciaran Connell about various technologies that can be used to determine location indoors, from Bluetooth to ultra-wideband (UWB).
Wong: Has location positioning been solved by iBeacons, using Bluetooth Low Energy, and by Wi-Fi-based applications?
This file type includes high resolution graphics and schematics when applicable.
Connell: There are a lot of technologies out there that have handled indoor location to some extent. Bluetooth and Wi-Fi can be used to measure location position to within four to eight meters, depending on how it’s implemented. But Bluetooth- and Wi-Fi-based systems cannot do really accurate positioning, like to within 10 cm.
Wong: Why do Bluetooth and Wi-Fi have that limitation?
Connell: Bluetooth and Wi-Fi, and narrowband radio in general, were designed for high-bandwidth wireless communication, but were not designed with location measurement in mind. The electronic structure of their radio waves make it difficult to measure a precise start and stop of a transmission, and this is made a lot worse when there are real-world effects like multipath and reflections off walls. Because of this, Wi-Fi- and Bluetooth-based location systems measure signal strength and use it to approximate distance, which is inherently not going to be that accurate.
Wong: Why is ultra-wideband (UWB) radio different?
Connell: Ultra-wideband, on the other hand, transmits in very quick bursts, making it easy to measure the start and stop of each transmission. This type of burst transmission is also much more resilient to interference, since reflected or multipath effects will be received slightly after the main transmission, and will therefore not interfere with it.
Wong: What do device makers find unique about DecaWave’s UWB product?
Connell: DecaWave’s UWB product is a single low-power chip that supports both location measurement and wireless data transmission. We’ve designed the chip to integrate well into everything from cellphones to robots, from employee tags to smart watches and keychains.
Wong: Does it really matter to track location to a10-cm accuracy?
Connell: We have a lot of customers who say yes. If you’re in a hospital looking for a medical device, where every second counts, you want to know the exact location of the device, not to waste time checking the wrong cabinet or storage room. If you’re making a home robot, you want to know its location precisely enough so as to not bump on doorframes when moving from room to room. And if you’re tracking a customer around your supermarket, you don’t want to offer him a discount on the potato chips you think he’s standing in front of, when he’s in fact standing in front of the pickles one aisle over. These and many more location applications require knowing location with more accuracy then five to eight meters.
Wong: How do robots fit into the story?
Connell: As I mentioned a minute ago, controlling a home robot requires knowing where it is to a reasonable degree of precision. Five or eight meters accuracy simply isn’t enough. Obviously, robots with precise location awareness will still need sensors to be sure they don’t bump into people’s legs or other objects whose locations aren’t fixed. But the ability to sense the wall before running into it doesn’t help the robot plan its movement. Precise location tracking is the key.
Wong: We’ve been discussing wireless and location. What does this have to do with the Internet of Things?
Connell: This has everything to do with the Internet of Things! Most of the “things” in the Internet of Things, both sensors and real-world “things,” need to connect using wireless radio, and need to track their locations. In some cases, we’re talking about tags attached to keychains and wallets that enable them to be found. In other cases, we’re talking about sensors that report the state of the world in their vicinity, and it’s very helpful to have the sensors also know where they are so that they can report accordingly. In these and many other Internet of Things applications, it’s important to use a wireless technology that supports both wireless data transmissions and accurate location measurement.
Wong: What about so-called smart homes?
Connell: We’re seeing a lot of our customers interested in smart homes, and smart homes also need to measure location precisely. If you want to adjust a room’s lighting, artwork, or music based on who’s in the room, you need to measure precisely who is in the room, who’s just outside the room, or who’s in the next room on the other side of a wall. If you want to customize a display screen, you need to know who is right in front of that screen and who is further away. Five- or eight-meter accuracy just isn’t enough.
This file type includes high resolution graphics and schematics when applicable.
Wong: What should we expect in the upcoming few years?
Connell: There are going to be a ton of new location-aware and smart devices, appliances, and real-world things. Whether it’s our phones helping us find our keychains or wallets, our wallets helping us find our phones, or living-room lights knowing which music to play as we enter the room and which pictures to play on the digital picture frame on the wall, or hospitals and offices knowing which cabinet each employee is taking things from, or robots knowing precisely where they are as they do their jobs, more and more things will track and deal with location.
