I’ve spent a lot of time thinking about the Internet of Things recently. Well, I’ve spent the last seven years thinking about IoT. Seven years ago, I bought my first Wink Hub. That and a couple of bulbs from GE started me down a path of exploring the connected home and various IoT solutions. But it was another two years before I jumped into my first automation and IoT project for my employer. The project, for lighting control in a massive manufacturing plant, offered two choices: a Zigbee solution or an IP connected solution using Wi-Fi. After investigation and testing, I carefully steered the customer away from Zigbee to the wireless solution. My reason was simple. Zigbee wasn’t ready.
Now, here we are over five years later, and I’m spending a lot of time thinking about, discussing, testing, and now writing about wireless IoT devices. What’s changed? Would I tell the same customer the same thing today?
The answer is a lot has changed. But I would still push the customer towards Wi-Fi. As most of us know, change doesn’t equal product maturity. However, I firmly believe over the next five years, we’re going to see significant development of IoT wireless protocols. I expect that period will radically redefine what it means to be a wireless SME. Why the excitement?
To understand wireless protocols currently, look back 18 years at pre-802.11 standard Wi-Fi. Like then, the first challenge is the lack of standardization. Most IoT wireless protocols are built on IEEE 802.15.4. However, they’re all proprietary or non-standard in some form or other. Zigbee is the most well-known of the standards, but others bring their specific twist and licensing models. Aside from the challenge of non-standard protocols, the second challenge is the lack of tools to support the networks. While Zigbee uses 2.4Ghz, a band where we already have spectrum analyzers, most protocols use either 400MHz, 800MHz, or 900MHz, depending on the region. Spectrum analyzers for these bands are difficult to find and very expensive.
Further, planning tools like Ekahau Pro aren’t ready to plan for coverage outside of 2.4 and 5Ghz. I expect Ekahau is working towards implementing 6GHz, so I don’t expect to see those lower bands soon. iBwave can support the required bands already; however, to do so requires a costlier license intended for LTE providers. Finally, Wireshark has an 802.15.4 dissector, but many protocols are still not supported, and documentation about what should take place in closed protocols is virtually nonexistent. Without appropriate tools, wireless professionals cannot support or troubleshoot connectivity problems.
Despite my pessimism around the current environment, just like the pre-Wi-Fi days, all is not lost! First, we have a growing market demand for wirelessly connected products. These products are in use in many industrial, manufacturing, and enterprise environments, and their place in the market is expanding daily. Next, Amazon, Apple, Google, and Zigbee have formed “The Connected Home Over IP” working group. Their focus is to develop “an open-source approach for the development and implementation of a new, unified connectivity protocol.” While the name clearly states this protocol is designed for the home, from watching Wi-Fi, we’ve learned consumers can drive their employers to adopt new technologies in the workplace.
Finally, the industry is taking notice. CWNP is in the development phase of a certification track for wireless IoT protocols to mirror their current path for 802.11 Wi-Fi. Aruba announced at Mobility Field Day 4 that some of their Wi-Fi6 AP’s are Zigbee capable, and there’s backward compatibility of some 802.11ac Wave 2 AP’s with a USB dongle attachment. As more vendors offer this ability, designing a robust IoT network may be as simple as designing great Wi-Fi.
Will 2020 be the rise of the wireless machines? Probably not. But just like John Connor, we must prepare for them now. We have a huge advantage on those early days of pre-standard 802.11 Wi-Fi. Test devices are now cheap and easy to acquire. From Raspberry Pi HAT’s for LoRa and Zigbee, to Adafruit Feathers, we can quickly begin learning about their weak spots and prepare for the rise of the machines.