Jump to content

Jon Eggins

Registered
  • Content Count

    6
  • Joined

  • Last visited

  • Days Won

    1

Posts posted by Jon Eggins


  1. Hi Adam,

    Thanks for joining and commenting. It's always great to see people getting in to this space. We host year 10 students for work experience, and they are all over it. Bodes well for the future, but also a bit of a kick for more experienced guys to stay current.

    Re you questions. 

    "In the short term, what would be a good place to start to learn how to get data from a Thing to a cloud?"

    Yes - and vice-versa. I would suggest AWS as the simplest entry to the cloud space, and then initially getting a feel for comms transport via some basic PC or mobile device app. There are many options for comms and transport. Choose one that has library support for your mobile device and AWS and use it. Better to know one REALLY well than 10 superficially. Don't forget the downlink - you'll want to be able to actuate you mobile device app from the cloud (presumably via a web GUI).

    You could make your own thing, or buy a thing, but a mobile device is a highly capable and transparent dev tool for things. That said, you may want to branch out to more things, say BLE-connected to your app, and then sub-squently available on the IoT. Again, at the local wireless network level, there are many options. Choose the ones that are best supported and learn them from end to end. BLE and WiFi are easy entry points. WiFi is possibly the simplest but is power hungry(er).

    Re Image gathering, and AR, certainly it is a hugely cool topic. It's not my particular area although we do do image recognition stuff & triggering of abstracted logic based on that. Yes the video is also more data and processing hungry. If you had a good data plan to play with you could upload to the cloud and process there if you thing was not up to the task.

    Check out https://artoolkit.org/.

    Cheers! Jon

    • Like 1

  2. Hi Stuart,

    I very much enjoyed reading your post. It was great to feel your enthusiasm.

    A good starting point for cost-benfit is with the saving of labour costs (and its morally/ethically nice to think that roles could be upskilled to where humans can be adding value that tech can't easily add, eg pattern matching, strategic decision making etc - tech is great & more reliable than us when it comes for laborious/repetitive work).

    So, if you can tally staff infrastructure costs (in as much detail as you can get) from people/cars/petrol/insurance/accommodation/etc etc, plus rationalise the cost of tooling for manual measurements (and its life cycle), then it may just flow in terms of a commercial model. 

    Do be prepared to present options, and perhaps the lowest end option would have a migration path to add on extra capability. Ie, have a base option that gets a foot in the door, so to speak.

    I would also suggest two perspectives, one being retro-fit to an existing plant install, and one being a "green fields" install. It may well be that in some instance that green fields will be cheaper, as the instrumentation may be easier to install during plant installation?

    I'm not an expert in your application domain however and so what I mention here is generic with my perceptions of reality :-).

    Cheers! Jon.


  3. On 5 July 2016 at 4:23 PM, Andrew at MEA said:

    “IoT is like teenage sex:   

    Everyone talks about it, nobody really knows how to do it.

    Everyone thinks everyone else is doing it, so everyone claims they are doing it too.”

    Gold!!!

    In all seriousness, it's true - most people really don't know or understand what IoT is meant to be. They thing its about inserting a sensor into the Internet, and then seeing what happens.

     

×
×
  • Create New...