Internet of Things News of the Week, December 1 2017

Get this news summary in your inbox every Friday when you subscribe to the newsletter.

The letter N may not be part of the Alphabet much longer: Google is reportedly bringing Nest back into the fold to consolidate Google’s hardware efforts. Google split Nest off in 2015 as part of its Alphabet restructure, but it makes sense for Nest to be within Google itself. Having two separate hardware divisions isn’t ideal, particularly when there are common threads (ha!) between the two such as wireless protocols, machine learning techniques and potentially overlapping uses of Google Assistant. By bringing Nest within Google, Rick Osterloh’s hardware team can continue to push forward with cohesive branding, design and functionality between various Google products. And that just might help in Alphabet’s growing battle for the home against Amazon Echo and Alexa devices. (WSJ)

Amazon goes IoT crazy this week: Speaking of Amazon from an IoT perspective, the company held its AWS re:Invent this week and in its own words, announced “a slew of IOT services.” I’d agree with that assessment. We’ll be digging deeper into the news but here’s a quick summary of what’s new for AWS: A preview of AWS IoT 1-Click that calls services with a button press, the AWS IoT Device Management framework, AWS Device Defender for IoT security policy audits, AWS IoT Analytics, the Amazon FreeRTOS platform for IoT devices built on microcontrollers, and the AWS Greengrass ML Inference that brings machine learning to embedded devices, even when they’re not connected to the internet. It’s worth noting that FreeRTOS was created by Richard Barry, who joined Amazon last year. The company also launched a deep learning vision camera for developers, so I expect more Amazon devices with cameras along with third-party skills or apps that can take advantage of them. (Amazon)

Oh, Google’s got a smart camera too but for tinkerers: I’d be remiss if I didn’t mention the second AIY kit from Google. In May, Google launched the AIY Voice Kit as a way to build a small Google Home-like device out of cardboard, a Raspberry Pi and a few other components. Now it’s following up with the AIY Vision Kit using similar parts. Included in the software are a trio of TensorFlow-based neural network models to give your home built camera some smarts. (Google)

HPE and ABB equals industrial IoT: This summary of the IoT alliance between HPE and ABB provides a nice overview of their industrial IoT efforts as well as a vision of using data for both continuous operations and analysis in a cyclical fashion. (Guardian)

Designing for Centaurs?: For those not familiar, the mythical Centaur is half-human and half-horse, or a hybrid creature. And that’s the analogy in this thoughtful piece on future design for users. Instead of creating software or services for people, designers will be challenged to create things for a hybrid user comprised of both people and machines. The Waze example is a perfect: Since Waze is monitoring a network of drivers in real time, it may send some humans on a longer route so that the network as a whole isn’t negatively impacted. A good thought piece here. (Co.Design)

AI and ML will make our smart homes smarter: The Centaur design article reminds me of a post I wrote earlier this week because I posit that our smart homes aren’t really that smart. What’s needed is more use of the knowledge graph our IoT devices can create about our unique living habits and then surface suggestions or take action based on human behaviors. I think it’s a good read that gives a glimpse of the smart home future, but then again, I’m admittedly biased on this one. (StaceyOnIot)

Smart IoT security guidelines: I love that this article notes 90 percent of security professionals consider connected devices to be the biggest threat they face, but it’s concerning that 66 percent of them are unsure how many such devices are in their environment. Time to check the router logs, gang! Regardless, we can never read too much about how to better secure the IoT and this list is a fine example. I particularly like the one that’s easy to overlook: disabling UPnP or Universal Plug and Play on devices. Given a choice of simplicity or security, I’d go with the latter every single time. (Innovation Enterprise)

Nokia is bringing IoT to Africa: Africa isn’t a continent known for manufacturing and production compared to other world regions, so it makes sense that IoT efforts there are focused on different verticals. And that’s what Nokia is doing: bringing IoT in the form of smart cities, public safety, connected vehicles and digital health. These are the areas that will benefit a broader number of Africa’s residents and tie nicely into Nokia’s networking and imaging expertise. For example: Did you know that Nokia has a worldwide IoT network grid as a service or WING for short? Me neither! (Talk IoT)

One multi-radio to rule them all? Since we’re traveling around the world, let’s hop from Africa to New Zealand where Pycom has launched FiPy. This board provides total flexibility for IoT device makers because it supports WiFi, Bluetooth, LoRa, Sigfox and dual LTE-M (CAT M1 and NB-IoT). Pycom’s approach is smart because it can sell the product to various network operators, regardless of which IoT radio they plan to support. Based on the FiPy spec sheet, this small hardware works across a range of frequencies for all of the networks, so it’s truly a worldwide device. And the board supports the MicroPython framework for developers to build their services. FiPy is now shipping at a cost of €54.00 for those interested. (ComputerWorld)

IoT trends for 2018 to watch for: A panel of experts (not including us – what’s up with that?) provide their 17 IoT trends to follow next year and I can’t really find any that I completely disagree with. With this being said, looking into something like IOT Cores could give you an idea as to what the future holds when it comes to the technological world. The first one, for example, is about the use of blockchain in IoT; something I hope to brush up on in the new year. If I were forced to pick one of these trends as a “meh”, it would probably the rise and potential of smart clothing within the next year. Yes, this will become more a useful thing at some point, but this market is maturing far slower than others in this space. Right now, most of the smart clothing implementations are super specific, expensive and don’t really solve many problems that another convenient form factor (think smartwatches and wearable sensors) already manage. (TechJini)