Another IoT Platform In The (Blue)Mix

Many major companies (Intel, Oracle, Atmel, and IBM, for example) are competing to be the standard interconnect fabric for the Internet of Things. As a developer, it is hard to cut through the marketing hype and decide which platform is the best for you and your application. Luckily, there’s a plethora of projects on the web that showcase these frameworks. These project sites are an easy way to evaluate the strengths and weaknesses of IoT frameworks in practical applications without having to develop prototypes yourself.

[diyhacking], for example, posted a demo of using IBM’s Bluemix along with a Raspberry Pi, to do some simple home automation tasks. The project hardware is modest, using a PIR motion sensor and a relay to control an AC load. However, that’s good because it lets you focus on the Bluemix tools. The example client and server software is less than 200 lines of Python.

Bluemix looks like it has good integration with the Raspberry Pi and features a simulator so you can work without real hardware for development. Bluemix does offer a free plan (with limits), but the fee options may be a turn off to some IoT hackers.

It was surprising, though, to see a home automation relay (which, presumably, is carrying wall current) put on a solderless breadboard. Although there’s nothing wrong with that in theory, it is often a bad idea in practice.

That still won’t stop you from comparing this project to Thingspeak or any of the other platforms out there.

19 thoughts on “Another IoT Platform In The (Blue)Mix

  1. How dare they, when the only reason the internet works is because everyone is using the same, open protocols and standards….the American government should rule that IoT manufacturers must work together on a shared standard.

  2. IoT is going to languish unless someone makes a universal, non-commercial standard.

    Physical standards are competing HARD right now. BTLE, wifi, 802.15.4, 3G, 4G, GSM, LORA, BOPL, even some on the old pager network. There needs to be a physical-agnostic set of standards that provides meta-information about the link and the link capacity at every layer.

    OpenWSN might save the world, but it is under-funded and is only 15.4 facing as of right now.
    I remember back in 2007, when it was still being called “Smart Dust”. Then the IoT name was invented, but no one said anything about it for another 7 years, and suddenly it is *all* people can say.

    At the ESC that just happened, everyone was touting IoT solutions, some only at the top level, some claiming bottom-to-top solutions. Supposedly, ARM is going to come in swinging with mbed 2.0 (or 3.0? whichever).

    We are in an absolute nadir of IoT tech and solutions right now.

    1. Andrew, I don’t think that is the main issue. Because in the end all devices reach the internet level and there they could talk the same language, independent on the PHY they used to get there. What you then need is some way for the systems from different manufacturers to talk to each other and I see this through multiple 3rd party apps that have the ability to talk to a variety of devices: openhab, openremote, mr house, freedomotic just to name a few rather free ones.

      When smartphones started appearing in the business world regular people had no need for them until they realized that they could communicate through all means at all time and for a low cost. It solved the need for communication, a very important human need. IoT has not found such a need to address.

  3. I am probably the only person who feels this way on HaD, but I have no desire to make / buy / whatever more things on the IoJ (Internet of Junk). It is probably just my own myopia, but I don’t see the appeal of connecting more things to the network. Is there already some sort of killer app out there that makes it all worthwhile? Or is it just stuff like letting you remotely toggle your lamp or start the oven when you are stuck in traffic?

    (Or maybe I just watched too much BSG and have acquired an inherent mistrust for networks… ;-) )

      1. I have a LIFX bulb and absolutely love it. I struggle to make up in the mornings, so I can set (from my phone) a time for my light to start fading on. I can set this on the fly, and to happen before my phone’s alarm goes off.

        Some uses for IoT are gimmicks, not all.

        1. If there’s nothing to be gained from it: a new solution, knowledge, or a cheaper solution; there’s not much motivation to ‘hack’ it. You can do it because you want to, but the more useless it is the deeper into ‘art territory’ you go. The point [The Big One] is making is that in their experience, IoT is a solution without a problem.

  4. IoT is sorta flailing for a compelling direct consumer problem in need of solving, but for commercial applications it’s been happening for going on a couple decades. Now we’re just finally getting cheap radios with bidirectional channels at a price point that’s an enabler. (Vs. say, $100 per endpoint for RadioTronix or Linx stuff half a dozen years ago…)

  5. i’ve been playing around with Mosquitto, an open source MQTT broker http://mosquitto.org/. it’s really easy to use with a a client on a ESP8266 nodemcu using LUA and it’s simple to configure the server (broker) on a Raspberry Pi. i’m sure this isn’t the “one-size fit’s all” solution IoT has been waiting for but when it’s super easy bridging networks running different protocols who cares? what’s important to me is that the information it carries is encrypted and that i can run my own server independently.

    1. Totally correct. MQTT is a great way of collecting data from IoT devices. It’s quite easy to write a bit of python code to listen on an MQTT channel for data from an IoT sensor for example, and then relay that data to a number of different systems. You could feed the data into thingspeak, bluemix, and a number of other applications without having to put the load on your IoT device. Same thing for control in the other direction. Just means you only need to write one bit of application code for your thing (MQTT), then bridge it to whatever platform you happen to be evaluating.

      Also, it’s not obvious from the article or the python code (using IBMs module), but it’s using MQTT in the backend :D

      1. Sorry about being late to the party. It’s actually easy to write code in just about any language. I write a Perl and Python script to allow me to call MQTT from bash. I currently have a number of programs with MQTT in the middle.

        The big one its Misterhouse, while I need to do further work on the MQTT module to support login, recovery and reconnect, and SSL the basics do work

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.