IoT news of the week for March 25, 2022

Why isn’t 5G happening in factories yet? This week on the podcast, Kevin asked why we have seen so much hype but so few deployments around 5G for manufacturing. I blamed it on the fact that Release 17 of the 3GPP standard was just finalized, so we can finally get network equipment and end devices to the market that are optimized for manufacturing use cases. Here’s Phil Skipper, director of IoT strategy at Vodafone going into that, and all of the other reasons. It’s a good interview. (Enterprise IoT Insights) — Stacey Higginbotham

Amazon shares plans for Alexa to become even more helpful: Would you ask Alexa to help plan your vacation? Not today, but perhaps in the future the digital assistant could handle such a task. That’s the gist of this article in which an Amazon SVP and the head scientist for Alexa walks the reporter through Amazon’s plans for “ambient intelligence.” There’s nothing new here and nothing about the amount of computing and input data needed to make Alexa capable of this advancement, but if you want to see where Amazon wants us to think Alexa can go, it’s useful. (Fortune) — Stacey Higginbotham

Sensorbee chooses Piera System’s chip for more accurate air quality measurements: Sensorbee AB will put Piera’s chip on its outdoor air quality monitors for smart city deployments. I wrote about Piera in the past because its technology allows it to accurately measure air quality particles as opposed to estimating them, and this seems like an increasingly important thing to do given the health consequences of both indoor and outdoor pollution. Plus, I love new sensing technology. (Piera Systems) — Stacey Higginbotham

Lexmark adds edge AI functions to its Optra platform: Lexmark introduced its Optra software last year and this week it launched Optra AI Edge, which lets its customers run their AI models closer to where the data is generated. This reduces latency, can lower costs, and also can help with data privacy — all things enterprises are keen on. Optra is built from Lexmark’s decades of managing millions of connected printers for clients, and after talking to them, it’s clear they have an understanding of the basics needs for an IoT platform. With this launch, we’ll see how Lexmark handles more complexity. (Lexmark) — Stacey Higginbotham

Snap wants to read your mind: I’m still not sold on the promise of AR/VR, but even I have to admit this is pretty wild. Snap this week acquired NextMind for an undisclosed amount. What does NextMind bring to the table? The company produces a $400 headband that lets you control computing interfaces with your mind using machine learning. Aside from the Snap Spectacles, which got people a little more comfortable wearing camera-equipped glasses, Snap has made a string of acquisitions in this space. In the past year, it bought two companies related to AR display technology. (The Verge) — Kevin C. Tofel

Nvidia’s new robotic product is a big AI brain: There’s good news for robot makers that need more brain capacity in their hardware. At its Nvidia GTC event this week, Nvidia unveiled a $1,999 computing box that puts up to 275 trillion operations per second in a mechanical cranium. Yeah, 12-core ARM Cortex-A78AE processors paired with Nvidia’s GPUs will do that. This is the top-tier developer kit configuration. For those with a smaller budget, there are $399 and $1,599 options that are obviously less capable but still great for building a brain. (Engadget) — Kevin C. Tofel

TinyMobileRobots has marked up a million sports fields: Speaking of robots, have you heard of TinyMobileRobots out of Denmark? The company focuses on automating one thing, and does it really well. The little robots (they’re too big for me to call them “tiny”) can mark appropriate lines on roads and athletic fields. And this week, the company dropped chalk lines on its one-millionth field. That’s a big number, but you’re still probably asking how big of a deal that is. According to the company, its robots have eliminated 1.6 million hours of manual labor, equating to a cost savings of $102.5 million. And those aren’t tiny numbers. (TinyMobileRobots) — Kevin C. Tofel

Arm might be worth more than Nvidia was paying: It comes as no surprise that after Nvidia failed to acquire Arm, Arm decided to go public. What is a surprise, at least to me, is that Arm may be worth up to 50% more in an IPO than what Nvidia was hoping to pay for it. That’s the word on the street, anyway. Instead of the $40 billion it would have received from Nvidia, an Arm IPO might value the company at $60 billion. Regardless of the valuation, a public offering is in everyone’s best interest compared to a purchase. With its chip architecture designs used in thousands of products from so many connected product brands, all that power shouldn’t be centralized within a single chip designer like Nvidia. (Reuters) — Kevin C. Tofel

Yet another Helium LoRa device is coming soon: If you’ve been waiting for your very own Helium hotspot like I have, there’s now a new option. Pycom is offering one with a price of €275, or roughly $302. That’s less expensive than the €375 hotspot I ordered from Nebra more than a year ago, and I don’t even have it yet. Since Pycom is using Raspberry Pi boards for the compute power, I’m sure these will be limited and hard to get as well. The company says it plans to ship the first batch of its Pycom miners in June. In addition to earning Helium tokens by deploying this LoRa hotspot, you can also use it with certified tracking devices. Pycom is happy to sell you one for an additional fee. (Pycom) — Kevin C. Tofel

Want to customize LED light strings? There’s an app for that! Most smart home owners that have LED lights typically use the manufacturer’s mobile app to customize and control them. I don’t blame you if you fall into this camp. But what if you want to build your custom LED string project with an Arduino board? Well, now there is a simple low-code/no-code method for that. I stumbled onto a web app that makes it easy to pick your LED colors and layouts. You can build your addressable LED design in Google Sheets, paste the content into the LED Mapper, and it spits out the Arduino code that makes your lighting your own. (Hackaday) — Kevin C. Tofel