IoT news of the week for July 31, 2020

Cloudflare is building the next-gen architecture for the IoT (and edge computing in general): I last spoke with Cloudflare CEO Matthew Prince in October 2018, when the company launched its edge computing platform, Workers. This week, the company updated that offering with several announcements and one of the best architecture posts I’ve read in a long while. I’m going to get Prince to come on the show to talk about this in depth, but it’s worth your time to read his blog post on where the company thinks serverless computing is going and how it is addressing the needs of the next generation of applications and services (including IoT). The post addresses the business rationale for edge computing and how that is changing before explaining how Cloudflare is building the tech needed to meet the opportunity. (Cloudflare) — Stacey Higginbotham

Google’s Coral AI at the edge gains broader availability: Google Research debuted edge AI chips in 2018, an effort that has since morphed into something called Coral. Now Coral has advanced to bring a number of updates so you can more easily do your own AI edge projects, even with the TensorFlow-based Coral chip. This week, Google announced that the Edge TPU runtime is open source, with the code available for both Linux and Windows. This will let developers use non-supported ARMv7 and RISC-V chips with modified code for edge AI solutions. And what’s an edge AI project without some data models? Google has added BodyPix, a Google person-segmentation model for person recognition use cases and training that was previously only supported on TensorFlow.js. (Google Developer Blog) — Kevin C. Tofel

Could IoT sensors help save trees from climate change? Defra (the UK’s Department for Environment, Food and Rural Affairs), Forest Research, and Vodafone are bringing IoT to the forests. By using wirelessly connected sensors attached to trees, researchers can gather temperature, humidity, tree growth, and soil moisture metrics from a vast number of trees without having to continually visit each trunk. This is just a three-month trial, but I’d be keen to see what the data tells us about the trees as our environment continues to change. (Euronews) — Kevin C. Tofel

Form smart swimming goggles can use the GPS from your smartwatch: There’s good news if you plunked down $199 for Form’s connected goggles, or if you plan to. A software update released this week adds the ability to use the GPS in your smartwatch over Bluetooth so you can track swim distances in open water. Obviously you’ll need a GPS-enabled smartwatch to use this feature, but the list isn’t too sparse. The Garmin Forerunner 945, fēnix 6 Pro, fēnix 5 Plus, and Apple Watch Series 3, 4, and 5 are all supported. (The Verge) — Kevin C. Tofel

OpenSpace raises $15.9M to bring 360-degree video to construction sites: You might not be familiar with OpenSpace, but the company is becoming well known in the construction market. The startup raised nearly $16 million in Series B funding this week to expand its services. OpenSpace uses hard hats with 360-degree cameras to capture site details and then sends the imagery to the cloud for processing and stitching. The end result? A happy construction manager that can review the site and its changes during the build process. Clearly there’s a market for this, as the company has reported a 500% increase in year-over-year revenue. (CrunchBase) — Kevin C. Tofel

Amazon lands a New Zealand IoT energy project: Amazon’s AWS has scored its first strategic alliance in New Zealand with Vector to jointly develop the New Energy Platform (NEP) for that country’s energy industry. The first step will be to collect and analyze data from more than 1.6 million IoT-connected Vector advanced meters, which will then be analyzed by the AWS IoT Analytics platform. The goal is to help plan energy networks, drive smarter investment decisions, and increase reliability for consumers based on smart meter data. (Vector) — Kevin C. Tofel

Wyze Cam Outdoor Camera review — a great device for the price: I took the upcoming outdoor camera from Wyze for a spin over the past few weeks and came away impressed. That seems to be a trend with Wyze devices; they work well and they’re a great value due to the low price. I liked the new Travel Mode feature, which lets you take your battery-powered camera on trips, as well as the video quality of this weather-resistant camera. At $39.95 a camera, is it an easy purchase decision? You might want to read about the glitch I experienced before pulling out your wallet. (Stacey On IoT) — Kevin C. Tofel

Beetles with Bluetooth are the newest webcams: Speaking of cameras, you can’t buy this one, but it’s still interesting. Researchers at the University of Washington created super-small webcams with tiny Bluetooth boards and attached them to beetles. Yes, beetles. As in, bugs. While it’s neat to see the world from a beetle’s point of view, I’m more amazed at the video quality and 100-meter-plus range over Bluetooth on these little science projects, each of which weigh a scant 248 milligrams (PopSci) — Kevin C. Tofel

There’s a new open-source IIoT edge project hitting the gas pedal: A slew of partners, including Dianomic, Flir, Google, and Nokia have helped the Fledge edge project move to the growth stage. Fledge was moved into The Linux Foundation and has matured from the “seed stage” with customer use cases, such as one from Gradient Racing. The race team used IoT-based digital twins of each track it competes on, a machine simulator, and an operator simulator to break two track records in the GT3 season last year. If you’re looking for IIoT solutions, this is one to watch. (LF Edge) — Kevin C. Tofel

Edgeworx debuts a thermal camera for temperature screenings at schools. It seems like every connected product is positioning itself as a helper during the COVID-19 pandemic. Edgeworx is the latest entry with Darcy, the company’s new AI-powered thermal camera. The big deal here is that high-precision thermal screening solutions, which can cost well into five figures, can take up to an hour for hardware calibration. Darcy uses software and sensor readings around the room to calibrate in just 5 minutes, and it can do so at a much lower cost. (Security Informed) — Kevin C. Tofel

Reach Stacey’s IoT audience by sponsoring this newsletter. Please email andrew@staceyoniot.com for a media kit.