• DIY for future reference.

  • TL;DR (you should read it though): Massive expansion into robotics and maybe household electronics.

    Dyson is aiming for a bold and potentially risky expansion of his brand, which some employees have taken to calling “Dyson 2.0.” Last year, the company broke ground on a more than $400 million technology campus adjacent to the Malmesbury headquarters. When it is completed next year, it will house 3,000 designers and engineers. Already, the company has brought in hundreds of software and computer hardware specialists and tripled the size of its engineering staff. The company currently funnels $2.5 million into R&D every week; it has spent $8 million on a cutting-edge robotics research lab at Imperial College, in London, and put another $15 million in the American-based solid-state battery company Sakti3.

    In coming months, the technology campus will serve as a launching pad for a range of new verticals, some of which Dyson has disclosed (robotics), some of which seem imminent (the Sakti3 investment appears to indicate a further interest in household electronics), and some of which are entirely classified.

  • Notes: 3

    Tags: Funny

    image: Download

    Ladies and Gentlemen, this is the Internet of Things (today).

    Ladies and Gentlemen, this is the Internet of Things (today).

  • “Creating a connected building in under 4 hours”. If you don’t include planning time! More strictly speaking, “Plugging in and taping on 1000 sensors in 4 hours”.

    One cool application is taping motion sensors to chairs in conference rooms so you can tell whether they are being used or not.

  • It supports a built-in TCP/IP stack, TLS and other various built-in security features (needed for secure actuators and sensors). But horribly, they did not announce a price. $20, $80, $120? Who knows? Announced for $USD 50.

  • TL;DR by quantifying risk by measuring real data.

  • TechCrunch article. The first part summarizes the problem space - oddly missing that the big problem with startup IoT products is that they’re likely not to be in business in a couple of years; after that standards that might be useful.

    Just to reiterate my thoughts: 

    • Make connected devices, not smart ones (less to hack)
    • Defence in depth - they couldn’t hack Stacey Higginbotham’s house until she gave up her WiFi password
    • Encryption - particularly TLS
    • End to end security - this is encryption again, but in particular the cloud is an excellent vector for hacking people’s Things
  • Goodbye Wheezy, hello Jessie. For me this doesn’t really make any difference because I’m all command line.

  • TL;DR all over the place / tried to be all things to all people. 

    Let me add a personal aside that they blew it on the IoT front by making Wink a hacker / hobbyist hostile box.

  • Nothing surprising in the list, and surprisingly high level.

  • TL;DR Amazon Echo, Philips Hue, Nest, Node+ ChromaPro.

  • Presence is a huge deal - see the tag “presence” to the right. Still, I think it has to be done with sensors independent of other Things.

  • They won’t change everything, unless there’s some amazing level of efficiency (and efficiency / energy-density / cost are the issues with solar), but I bet there are some cool applications, especially with low power sensors.

  • This seems to be more about interoperability and using sensor data for spying.

    We live in a ‘sensor-rich’ environment. These sensors connect in ad-hoc, local and mesh networks, operate using different protocols and standards, and use encryption to protect communications. Through this competition we want to know:

    • how can we pull together and visualise data on subjects of interest from a diverse, sensor-rich, loosely connected environment that operates across different protocols and across standards?
    • how can we show how sure we are of a hypothesis that 2 events/users/devices are connected, based on the data gathered?
  • Nest has released an API for it’s Camera, available through their Firebase, REST and REST Streaming APIs. The Nest Home Simulator app also now has a camera simulator. Cool.

    I’ll be updating HomeStar Nest to support this - in case you want a straight forward NodeJS way of accessing it. Nest also has a Node API here.