IoT news of the week for June 9, 2023

This company wants to launch a satellite-based Bluetooth network: Yes, this story initially seems bonkers, mostly because Bluetooth is a short-range technology that couldn’t possibly work from space. But Seattle’s Hubble Networks wants to build a constellation of 300 satellites with technology that essentially “magnifies” Bluetooth signals to let them travel up to 1,000 kilometers (621.4 miles). The company will launch four satellites first to test out the concept. Look, I’m as excited by Bluetooth as the next person, but I don’t know if we need yet another satellite network floating above us to provide a network that we can already create with phones and specialized tags today. (Hubble Networks)

A star invests in satellite sensing: Nuview, another startup with its eyes on the sky, has raised up to $15 million for its Lidar sensing network, with some of the funding coming from actor Leonardo DiCaprio. The company plans to generate an annual map of the Earth’s surface that provides centimeter-level accuracy. That sort of accuracy could provide important insights for precision agriculture (companies already prefer to use satellite imagery over soil sensing in some parts of the world) as well as deeper insights into how the environment is changing over time. If the company had the ability to quickly generate such precise maps, it would be a huge boon to the defense and surveillance industry. (TechCrunch)

Lidar sensors are making their way to power lines: While Nuview is investing in Lidar in space, AES, a utility, is investing in Lidar sensors to track the health of its power lines. AES has signed a deal to test no-touch Lidar sensors from LineVision Inc. on its transmission lines. The pilot attaches the LineVision gear to a transmission tower and uses Lidar to detect the health and quality of the transmission lines above. With the health data, AES can determine whether or not it can flow more power over those lines, enabling more electrical capacity without investing in more physical infrastructure. This is one way of using sensing tech to push more power onto the grid when necessary. (LineVision)

Arduino’s Pro business for industrial IoT is growing: Arduino, which makes an open source hardware platform for controlling physical devices, has expanded in the U.S. and hired an IIoT expert to lead its growth in the Pro line of business. Arduino notes that 39% of Pro sales have been in the U.S. and sales have doubled each year since launching the line in 2020. So the company is opening sales offices in Austin and Chicago, and has hired Guneet Bedi as senior vice president and general manager of the Americas. I spoke to Bedi in his roles at Cisco and at industrial IoT startup Relayr; he’s a good person to lead the sales efforts to bring Arduino gear into the industrial and medical device worlds. (Arduino)

Get ready for some more fog computing jargon: When it comes to cloud computing vs. edge computing, the edge appears to be winning. But it’s not a zero-sum game, and especially when it comes to using AI, the emphasis will be on sharing data and models in both places. Broadly speaking, training will continue to happen in the cloud, while inference will occur at the edge. This is leading technology vendors to come up with new jargon to describe sharing data between local and cloud. Qualcomm calls it hybrid AI, and others use the phrase Edge AI. These are fine. But over in the EU someone has created the term Cognifog (cognitive + fog) to describe this future of computing. The project using this jargon just scored €5 million ($5.4 million) in research funding. Ugh. (Qualcomm, EU)

Talk to your robot with ChatGPT: I really liked this article on how generative AI, including ChatGPT and DALL-E, might change the field of robots. The most obvious way it could help is by letting us talk to robots as opposed to programming them, but it also covers how using generative AI models associated with images could create synthetic data that could help train robots. It covers many other positives, and even includes a fairly nuanced discussion of some of the drawbacks, including the cost of training models for use in robots. Go read it. (The Robot Report)

We need new security standards for IoT sensors and to implement what we already have: Nothing in this article is new, but it does provide a good overview of the existing standards available for IoT sensors used in industrial settings and explains why these standards aren’t always implemented. It also discusses the expanding attack surface and the risks of having connected sensors throughout an organization. If some part of your role requires you to know something about securing operational technology, you should read this. (Communications of the ACM)

WiTTRA gets FCC approval for its LPWAN gateway: WiTTRA is a Swedish company building massive IoT networks using a 6LoWPAN mesh network with a mioty LPWAN radio. It is one of dozens of companies pushing proprietary forms of Low-Power Wide-Area Networks designed to send small amounts of data from thousands of sensors over relatively long distances. The mioty technology was developed by the Fraunhofer Institute and has a few adherents. It’s still fairly early for wide-scale corporate sensor networks that comprise tens of thousands of sensors as opposed to hundreds, though. When the industry gets to that multi-thousand-node scale, it’s possible that existing LPWANs will become less economical or tough to use without interference, battery changes, or other issues. So I’m keeping an eye on technology deployments using these alternatives. (WiTTRA)

Atmosic chips win a place on Google’s TV remote reference designs: Atmosic, a company that makes a Bluetooth radio and MCU that runs on harvested energy, has seen its system on a chip incorporated into the reference designs for Google TV remote controls. With this design, such remote controls could be built to operate TVs using buttons and voice, all without needing a battery. The reference designs are part of the Android Open Source Project (AOSP), which provides an open platform for companies to design products that are compatible with the Android operating system. (Atmosic)

Aqara has launched its U100 smart lock in. the U.S: Chinese smart home device maker Aqara has launched the new U100 Smart Lock in the U.S. The device costs $189.99 and lets folks unlock their door using keypad, fingerprint or Apple’s Home Key. The lock is compatible with Apple’s HomeKit, Matter, Google Home, Alexa, IFTTT, and more.  The lock has a Bluetooth and Zigbee radio, and for some of the features to work (like Matter) you’ll need a Matter-compliant Aqara hub. (Aqara)