IoT news of the week for July 19, 2019

Tuya inks a deal with Soracom: The Chinese company that provides the cloud back-end of more than 100,000 IoT companies has signed a deal with Soracom, which allows companies to pay for networking and connectivity on demand. The deal will offer each company’s services to the client base of the other. These sorts of partnerships make sense as most companies don’t want to build and integrate their own internet-connected devices. Many would rather focus on their core business and let others provide the nuts and bolts of cloud support, connectivity, and even app development. (PR Newswire)

AT&T and Microsoft team up for buzzword bingo: This news release has it all! There’s AI, cloud, and 5G! But what does it mean that Microsoft and AT&T are “embarking on an extensive, multiyear alliance where the two companies will apply technologies, including cloud, AI, and 5G, to improve how people live and work today and in the future?” The jargon is hiding a simple announcement: Microsoft will become cloud provider for all of AT&T’s non-network-related computing, which is a deal that should have happened years ago except AT&T really thought it could become a cloud company. And Microsoft will test and build products for AT&T’s 5G network. But it was going to do that anyway. What would be interesting (for AT&T) is if Microsoft was going to exclusively partner with Ma Bell to build edge processing into AT&T base stations or develop products for the cell phone network edge specific to AT&T. (Microsoft)

I’m not buying this: Pampers has introduced Lumi, a line of products that features a Logitech camera that will track your baby’s diapers, movement, sleep, and more. This isn’t the best idea, not only because it will be expensive and is largely overkill, but because Wi-Fi-based systems are vulnerable to internet outages and power outages. Witness how when the folks using Nest cameras as a baby monitor get mad every time Nest’s servers go down. The most reliable baby monitors will connect directly to a monitor in your room over a proprietary frequency. As for the benefits of tracking your baby’s wet diapers, your baby will let you know when they need a change. In general, despite what new parents think, babies are resilient, so unless yours has a medical issue this system seems like overkill. (Engadget)

Real talk about smart cities: This opinion piece hits all the main points as to why we should question the adoption of smart cities. The concerns the author has about building smart cities applies equally to smart devices in the consumer home. Not only do you make things more complex and offer a wider attack surface when connecting cities or homes, but you also introduce elements that will need more updating and will cause privacy concerns. I’m not against smart cities, but I am all for applying technology in ways that make sense, and only when the risks of technology are outweighed by the new service. (NYT)

Micron’s deep dive into sensors and AI to boost production quality: This week you get two stories about memory firms using AI to improve quality. In the podcast (above) I interview a Seagate executive about how the company applies computer vision to improve yields, and in this story, the author writes about Micron’s efforts in the same area. The article also talks about the challenges Micron has had trying to add acoustical analysis to its factories. (CIO Dive)

Prefer your computer vision stories with less tech and more tomatoes?Domino’s is using a scanner to ensure that its pizzas go out the door looking pretty, with the right number of toppings and the proper disbursement of cheese. The company lets consumers view their pizza as it’s being cut as well. I’m not likely to check in on that, but it’s nice to know I might not get a pizza that has only four pepperoni pieces clustered onto one slice because the teenager making the pie wasn’t paying attention. (Future Five)

Mnubo has been acquired by AspenTech: Mnubo, a Canadian data analytics company, has been acquired by AspenTech, which provides asset optimization software. The deal was valued at $102 million CAD ($78.3 million) AspenTech also announced the acquisition of Sabisu Ltd., a UK firm that specializes in data visualization software. I first encountered AspenTech a few months back at an event and I should definitely find time to chat with this company. It has come up recently in several conversations with IIoT folks. (CBC)

This will be how AI takes jobs: This story about one truck driver steering two trucks on a closed track is an example of how we’ll see AI “come for people’s jobs.” Autonomous vehicles are not so autonomous that they don’t need the occasional human brain to handle a complex scenario. So what we’ll see is a gradual scaling of the number of vehicles that one human can pilot. Today it’s a 1-to-1 relationship. In the next five years or so we may get to five trucks per human. It’s similar to the scaling that happened in the data center with sysadmins and servers. It started with a 1-to-5 ratio and has since moved to 1-to-1000 or more. So job loss will be gradual (for example, today we don’t have the long-haul truckers needed for all routes) and then happen all at once. Except it’s likely that having robots perform mundane tasks really will open up new forms of work for people. The key will be training and retaining workers as that happens. (IEEE Spectrum)

ChefSteps is acquired by Breville: I have a Joule sous vide cooker made by ChefSteps, and I like it a lot. I’m just not super into sous vide cooking, which means I don’t use it very often. I’ve also been worried about the longevity of the company and its app, especially given that the Joule doesn’t have any physical controllers. I may have been right to worry. ChefSteps was acquired by Breville this week in a fire sale, which means that I don’t know what will happen to my $200 kitchen gadget. Based on this interview it sounds like I should be good for a while. (The Spoon)

Here’s my review of the Wyze connected bulbs: They are cheap, good, and worth purchasing. (StaceyonIoT)