Running a wireless network in an industrial factory is tough. Not only are there large hunks of metal that move around, blocking signals willy-nilly, but there are randomly timed events that generate electromagnetic field interference. If your sensors are trying to send essential data, such uncertainty can be debilitating.
Manufacturers just can’t work like that, which is why there are tons of proprietary wireless protocols for manufacturing clients. But this multiplicity of protocols also makes it tough for companies to take advantage of less expensive commercial-off-the-shelf networks. Austin, Texas-based Spearix hopes to solve this problem with a new chip that can tackle multiple wireless networks smoothly and on the fly.
Manas Behera, CEO at Spearix, told me his goal was to find a physical solution to the wireless problem in factories, not just some kind of software wrapper. To that end, Spearix is designing a new chip that can hop easily between frequencies and protocols. So far the company can support 802.15.4 radios such as Thread, WirelessHART, and Zigbee as well as Wi-Fi.
The company has built a multicore processor that will handle different radio frequencies with a processor dedicated to understanding which core is performing best at any given moment. Each radio core can wake up to try to send or receive data using a different frequency and protocol depending on what the situation calls for at the time. So if Wi-Fi works best, then the Wi-Fi radio will send data, but if that stops working the chip will switch over to a Zigbee or WirelessHART core and start sending.
The workhorse of the chip is a processor running what Spearix calls its RADiS (RF Adaptive Diversity in Synchronized networks) technology. The processor analyzes both the environment and the quality of the signals and performs processing on all incoming signals to select the best network for the job. The goal is to focus on sending data reliably and without high-power consumption.
Reducing power consumption is another reason for the multiple cores, so when they aren’t sending data, they are not using power. Behera explained that the technology can be used in various different spectrum bands and for different protocols. Sending 80 bytes of data costs about 100 microwatts of power, which means that, depending on the protocol, transmitting once per second could enable a 10-year battery life with an AA battery.
This is incredibly efficient and becomes even more so if both the sending radio and the receiving radio are using the Spearix technology. But Behera is starting by putting the chips in gateway devices, where they will still offer some power and performance savings while not requiring customers to get rid of their older equipment.
This makes sense, and is likely a reason that Emerson has invested in Spearix. Emerson, which has pioneered the WirelessHART radio technology, is well aware of the challenges industrial environments pose for wireless data transmission, and is piloting the technology. Behera told me that the National Institute of Standards and Technology is testing the tech as well. He expects the first module to come out in early spring and is looking for others to test it, too.
As a chip nerd who loves digging in on wireless technology, I’m excited about what Spearix has on hand. It could work to help manufacturers that are investing in a variety of radios, especially those that may not be widespread, such as private 5G over the Citizens Broadband Radio Service band spectrum.
Today, the tech would likely compliment an existing chip company’s portfolio, making it a possible acquisition target or partner for a company such as Silicon Labs, Qualcomm, or Semtech. But it could eventually become a standalone business if it can successfully grow and find traction with large customers willing to take it beyond the pilot stage and into wide-scale production.