X
Home & Office

Zigbee inside the Mars Perseverance Mission and your smart home

ZDNet talks with Tobin Richardson, CEO of Zigbee Alliance, about wireless networking on Mars and the future of smart homes.

Have you been following the Perseverance rover that landed on Mars in February? It was fun to watch as the robot landed on the surface of Mars, ready to explore. And it's been even more mesmerizing to watch as videos -- with sound -- and pictures have made their way back to Earth. Part of the technology that makes the communication between the rover and NASA possible is Zigbee. 

My ZDNet Jason Squared co-host, Jason Cipriani, and I recently had the opportunity to interview Tobin Richardson, the CEO, and president of the Zigbee Alliance, about this project, as well as the future of the Zigbee wireless standard. 


Jason Cipriani: Tobin, thanks for joining us today. If you don't mind, tell our readers and listeners a little bit about yourself. 

Tobin Richardson: Tobin Richardson, CEO of the Zigbee Alliance. I've been with the organization for the better part of a decade and first joined Zigbee Alliance to help it get into smart meters around the planet. And then, as it became a more mature technology, I stayed on as a CEO to help the organization grow into a lot of different market segments, which is why you're seeing us in smart homes, smart buildings, industrial automation, and its use on the Perseverance Mars mission.

Jason Perlow: We have occasionally discussed Zigbee and other wireless data communications technologies used in the home automation industry and other verticals such as wireless sensors and industrial control systems. For our listeners who may not be familiar with it, can you tell us a bit about the Zigbee standard and the typical use cases? 

Tobin Richardson: It started almost two decades ago, and I had some experience with Wi-Fi and Bluetooth early on when they were more on the proprietary side. Those two have excellent use cases, but the use cases that drove Zigbee early on were around Personal Area Networking (PANs) or industrial wireless sensor controls and networking. This was based on an IEEE 802.15.4 standard, specifying how to implement that standard with what we call the Zigbee Stack. And early on, that was really about lighting systems, industrial controls, and wireless sensor control networks. And that was the first area where it entered the marketplace.

Jason Perlow: Today, the Zigbee protocol has a maximum transmission rate of 250kbps. That's significantly slower than other low-power data communications protocols like Bluetooth Low Energy which caps out at about 2Mbps. I understand that there are important differences between the two in how they perform and what situations you might choose one over the other, and what distances they are effective at. Do you see them as complementary technologies? 

Feature Set

Bluetooth

Zigbee

Frequency Operation

2.4Ghz-2.483Ghz

2.4Ghz and 900MHz

RF Channels

79

16

Modulation

GFSK

BPSK/QPSK UWB

Cell Nodes

8

65,000

Bandwidth/Transmission Rate

2Mbps

256Kbps

Range

10 Meters

> 100M using 2.4Ghz, 1km with Sub-Gigahertz

IEEE Standard

802.15.1

802.15.4

Tobin Richardson: So, as an organization, we have a lot of different technologies and applications. The number of technologies we have and how we relate to Bluetooth and Wi-Fi use cases is growing. Zigbee can have up to thousands of nodes and is a much longer range. There are other use cases where there are one or two devices, and it's OK to use the Wi-Fi or Bluetooth standard for those. For Zigbee, we've seen it grow for lighting applications, especially if you are at the San Diego Convention Center. At one point, I think virtually all the lighting there was done with ZigBee; you've got hundreds and hundreds of devices off a single network controller. So it's perfect for really large, diverse networks, and the range is still pretty impressive. 

On Zigbee, you're going to get 50 meters using a Smart Home application. On Bluetooth, you'll experience drop-off on your headphones or speaker after five or ten meters, depending on your specific use case. Bluetooth is doing some good work. And a lot of these standards will evolve. And we'd love to see what Bluetooth and what Wi-Fi is doing in a lot of complementary areas and where Zigbee continues to evolve. The original Zigbee stack is on revision 23. So we keep growing, exploring how that looks, how the technology functions, and it's got a really good sweet spot around diverse networks that take advantage of mesh.

Jason Perlow: What is the history of the protocol, and where are we at today with its feature set, with 3.0?

Tobin Richardson: It's been a fascinating kind of evolution. And honestly, one of the things that have kept me personally involved is how the standard is evolving. So if you think back to where it started, Wi-Fi, at that point, was really about the network -- so a typical use case is a laptop or a desktop and an access point, where you've got a limited number of devices. You're just really throwing packets over a network and gaining access to a web server, and for things like that, it's fantastic. Zigbee came in with kind of the same approach; we're just going to connect all these devices, we're going to figure out how the networking works. And then we'll just let people figure out what they're going to do with that, with one node connected to a controller or another node connected to 25 different nodes, let them route appropriately. So that's kind of the origin story, as you have these really lightweight communications. 

You mentioned the new 250Kbps transmission. When you're looking at the packet sizes and things like that, you're not going to be serving web pages over that. But also, you might build a light bulb; I might make a light bulb, and someone else a light switch. And if we're all doing that, in proprietary ways, all our on-off commands are different. And as you and your audience probably know, you can argue about on and off and what that looks like; you would think it's binary. However, it's not quite so much in terms of how you turn that into something that's implemented. Again, this is part of that evolution, where we started, it was about how to apply the IEEE 802.15.4 standard and choosing how best to do the networking, and getting into the application. 

So for Zigbee Alliance members, there were many lighting companies and a lot of building automation companies who effectively were doing things a little bit differently. However, they said, let's build a standard, let's agree on what that looks like, let's agree on what on and off looks like, let's agree on what kinds of currents we're going to use, as well. So this turned into a new area of work for the Alliance, but it was still tied together in one stack. And so you've got the Zigbee professional networking component, we started building this application layer on top. And that's really what's led us to where we are today, in the Zigbee stack. 

We still have a flexible mindset. But we could have a lot of applications in medicine -- cool. Let's go off and do that. We might have some really good energy applications, so let's define that application layer in energy. Super. How about home automation? Great, let's go to that. 

But in home automation and energy, both have thermostats. So are we going to define thermostats differently? And these separate application profiles, okay, we're defeating the purpose. So we brought that back together -- and that's what Zigbee 3.0 is today, which is a really good natural evolution, right? So today, 18 years later, starting with the lightweight personal era and industrial, you now have a full-stack Zigbee 3.0 standard that defines all this. And that's in our revision 22. We're working on our revision 23 right now, where we'll start getting into usability across various hubs. And it's all part of that kind of Zigbee evolution, to standardize as much as possible with that alignment of the nodes. So the controllers all work together seamlessly, with a consistent language for those devices that are attached as well. 

Jason Cipriani: What role is Zigbee playing with the Perseverance rover and the Ingenuity drone on Mars?

Tobin Richardson: I had the good fortune of being part of a retreat put on by Amazon five years ago, where I met the project lead for the Mars Rover and finding out about amazing work that they're doing, which we didn't even know about early back then. And I don't know if Zigbee was part of it. But this is about communication between the rover and the helicopter. The helicopter flies autonomously when it's up in flight, but it can transmit data back about the location or other information about battery and things like that when it lands. So it's the mission telemetry that can get back to it. You've got a Zigbee 900 megahertz radio on the rover and another one on the helicopter itself that can communicate that way.

Jason Cipriani: Why is Zigbee suitable for data communication between the Perseverance rover and the Ingenuity drone? What sort of data is being transmitted? 

Tobin Richardson: Zigbee is the way to go. And, and I will defer kindly to them and let them explain that, but from my perspective, I think we're making a lot of sense for them as the low power component to this with the low data rate. Looking at really extreme environments, like Mars, it's good to have a very lightweight purpose-built standard. So it was built up from that perspective, where it's essential to get the basic information across and makes it possible for extended battery usage for those kinds of applications on Mars. I haven't seen a power system up there yet -- I don't think Matt Damon's put in one yet, as far as I know. I believe that that low-power component makes it a really attractive solution for that application as well as the sub-gigahertz frequency range for longer distance communications.

1600px-anatomy-of-the-mars-helicopter.png

Anatomy of the Ingenuity Mars Semi-Autonomous Drone

NASA JPL

Jason Perlow: From my understanding of how the Ingenuity drone works, it's semi-autonomous; it's not a fully intelligent thing. It's more like a ride that you would see in an amusement park running down the track -- an invisible telerobotically scripted, pre-programmed route that uses telemetry that will be sent between the rover to the helicopter. As I understand, that track can be adjusted on the fly as needed, but on Mars, there is no Global Positioning System, so any positioning and navigation are being done with cameras and sensors. A lot of telemetry signaling occurs between the drone and the rover and then back to JPL through the four Mars satellites. Zigbee at 900Mhz has a maximum effective distance of about 3000 feet, so that's within the mission profile of what the helicopter is doing. You're not going to want to do a 3000 foot Wi-Fi transmission or even a Bluetooth connection. I can barely get Bluetooth to work 15 feet away from my desk, let alone 3000 feet.

Tobin Richardson: These technologies have great use cases, right? And no, not at all, not good Bluetooth or Wi-Fi use cases. This is not the right application for that. There are a lot of challenges in the operating environment too. We were talking about this on the team as this became public what other real-world cases there are where this might be useful. I'm not going to say there are Mars-like environments on Earth, but there are places where it's difficult, and you need high reliability -- remote areas that don't have access to a lot of the power capabilities in just a typical building. In places like pipelines and other remote areas, where you want to get good telemetry and want something that you can rely on, there are many good use cases there. And yes, Mars, this is one of those use cases.

We're fascinated by what's happening over this implementation. I think there are a lot of areas really in power usages, such as the transmit power and the transmission rates, and getting a better understanding of how that operates, in negative 40-50 degrees Celsius environments, we're really very curious about how that works and in terms of what we might learn from that, as well as packet delivery failure. Zigbee is really good for that in terms of retries and things like that. But those are a few of the areas that we think would be really interesting to learn from. Of course, this is a demonstration project, the way that NASA JPL has described it, they've set the expectation that this is the first time they're trying, so they've already learned a lot in terms of the data. We certainly hope that they can get good separation, get the missions and the flights to do they're expecting to, and get some good learning from them. 

Jason Perlow: Is Zigbee involved with any of these emergency field worker apps, like text device capabilities -- like potentially putting a Zigbee chipset inside a smartphone? So, for example, If a 4G or 5G network infrastructure were to go down in an emergency situation, would it be possible to do mesh network texting and maybe some rudimentary burst voice capabilities between handhelds?

Tobin Richardson: You know, you should be able to do that, but I'm not familiar with these directly. I know there are organizations like FirstNet that are looking to serve first responders as well. It's happening with fire departments; those are the things you're talking about, right? There are areas in which the technology is being used in new ways, such as in those field environments, such as where you're dropping sensors to track where the fire line is in a wildfire. Certainly, from a human perspective, tracking people in distress is instrumental in positioning emergency signals. So certainly, those are areas that Zigbee can be used. And, and as we evolve as an organization, there are other technologies we have in our house, with this common language for devices that we think can be used across technology. So not just a Zigbee network, not just a narrowband IoT, or 5G, but you can do a mix of those together and effectively have one common language kind of going across those different mediums.

Jason Perlow: A lot has been discussed about Amazon's new mesh network, Sidewalk, for use in its Echo smart speaker devices, which is implemented over its built-in Zigbee transceiver modules. Zigbee has been designed to be secure so that it may operate over private networks and not interfere with or cross-traffic with other nearby Zigbee networks. Amazon has altered the use case by having all of their Echos, regardless of who owns them, communicate over Sidewalk to share firmware patches and such. What is Zigbee's position on this? Do you feel there is a good use case scenario for public mesh networking with Zigbee outside of Amazon's Sidewalk?

Tobin Richardson: That's an interesting question; I think we're still kind of in a wait-and-see on Sidewalk and see where Amazon goes. Amazon is very active in the Zigbee Alliance. In fact, they're on our board of directors; we have some terrific engineers and principal architects that participate both in the Zigbee side and Project CHIP (Connected Home over IP) and in the MACfi stuff that we do within the Zigbee Alliance. Having a little bit of latency, I think in terms of the public networks, the way Amazon is doing it, that's a fascinating approach. There are some areas that we want to look at a little more in dealing with privacy and security. And as you said, in terms of how secure this is, how the mixed networks operate together. And that's an area that we want to investigate a little bit more, let's say for now, but right now, a little bit of wait and see on Sidewalk.

Jason Cipriani: Narrowing down more on Zigbee, how does the relatively new IoT Thread protocol compare? I say new, only in that consumers can finally start using it with Apple's HomePod mini and some accessories. 

Tobin Richardson: In terms of low power mesh networking, we really kind of are sitting in the same area. The Zigbee network is not native IP, necessarily.  Of course, you can easily map to it, and you can address a single device on a MAC address and things like that. So it's not without addressing, but the notion of IP has been one that we've been tracking closely over the 20-year history of the organization. And Thread came around right about the time as an organization we were looking at developing a similar low power IP stack with a lot of the same functionality. When we learned about that, with our sister organization, we said, do we create a competing one, or do we partner with them, because we knew that that language is going to operate not just on Zigbee networks, but also on other IP networks. So we decided at that point that we would partner with Thread Group. And we've contributed quite a bit to their development as well, even on the McAfee side of Thread. And so we're effectively a good sister organization with Thread Group. As they get to commercial rollout, we will have an application layer on that and Project CHIP. The differences today between Zigbee and Thread today are mainly around the IP addressability, probably the most known difference between the two. But we see a lot of synergies there with the organizations. Today, if you want to build a quickly usable product and in virtually every ecosystem on the planet, Zigbee is a great choice for you. As you look at this evolutionary piece, we kind of see this competence in terms of IP with Thread as a fantastic solution. And, and we think the right language and the right standard on top of that will be Project CHIP on top of Thread. And that will have a lot in common with what we do on the ZigBee side and the development side will be a lot easier there as well. 

Jason Perlow: Is there a Zigbee 4.0 in the planning stages yet? What improvements can we expect to see from Zigbee in the future? Have we improved data rates?

Tobin Richardson: Zigbee 3.0 is kind of how we're describing the complete stack. We did that when we brought the different profiles together. And so we're continuing on that path. Right now, we're working on our revision 23. One of the biggest functionalities in R23 is focused on what we call "All Hubs." And that's effectively trying to get all the hub operators to effectively treat devices with the same route joining processes and other pieces. So there's a lot of good improved functionality for consumers. Hopefully, consumers will just enjoy it in a cleaner, crisp experience getting devices into the network, regardless of which hub or devices they use. Also, we're going to be adding some support for sub-gigahertz in R23. So we'll start taking advantage of other channels and frequencies. And we've had some demand for that in different markets and market segments, whether it's home automation and smart energy, as utility companies want to try and reach larger places. Sub-gigahertz and 2.4 gigahertz have different behaviors regarding how they act with interference and barriers, and sub-gigahertz in big thick concrete buildings is a nice solution. In the UK, we've actually already done that. So we'll bring that over into our R23. So I don't think we're going to be calling it anything different, but there will be more functionality in the next release, which should be a really good improvement for consumer experiences on smart homes. 

Jason Perlow: Thanks, Tobin. Looking forward to everything that Zigbee is doing on Earth and other planets.

Editorial standards