There are millions of devices and sensors connected to the Internet, and the next decade will bring billions more. How will anyone keep track of all these sensors? With analog.io, a platform for IoT devices, and [Luke]’s entry for The Hackaday Prize.
The problem of aggregating data from an Internet of things has been tackled before. Last year, Sparkfun released data.sparkfun.com, built on Phant, a tool for collecting data from the Internet of Things. Even though Phant can collect the data, it only does this in neat columns with values and time stamps. To turn this into something a little more visual, analog.io was born. In the future, [Luke] will add support for thingspeak and Xively data streams; the entire project is intended to be backend agnostic, allowing anyone to get their data from any thing, store it on any server, and connect it to analog.io for visualization and sharing.
Graphing data provides for some interesting opportunities, like when [Luke] found his Internet-connected water meter was logging far, far too much water consumption. A fitting on a garden hose came loose, and the hose started pouring water onto the ground, a foot away from his basement wall. That’s a swimming pool’s worth of water on [Luke]’s foundation, easily and readily graphed. He’s now adding an alert feature to analog.io.
Graphing data does present its own problems, like when a sensor sends a single erroneous data point. [Luke] is calling this a ‘burr’, and analog.io can filter out these small spikes that make data unreadable as a graph. There’s a lot of work that goes into making a usable graph, and [Luke] is crossing all his ‘t’s and dotting all his lowercase ‘j’s.
While many of the entries for the Hackaday Prize are running at the ground level with individual sensors connected to the Internet, [Luke]’s project tackles the Internet of Things problem from the other end, providing everyone a way to easily visualize their data. It’s a great Hackaday Prize entry, and will surely come in useful for a number of other prize entries as well.
Another website HADDOS’ed
So far so good but fingers crossed :)
nice, I was looking for something just like this after moving from xively to phant and google charts.
looks like he’s just placing the data into highcharts.. wouldn’t really call this a hack any idiot can do it!
You’re right, that is certainly a good part of what the application is doing right now. My larger goal create a community around the data and have more features for analytics and things like that.
My focus is really on easy of use, if you have a data.sparkfun.com public key that points to IoT data, you can have a highcharts plot of that in a few seconds.
Well, I’m not even an idiot then because I don’t know how to do it. So yeah, useful to me!
will the software available to be run at home? Mostly IoT data is not supposed to be publicly available for others and no, I don’t consider cloud solutions being private.
Why read when you can just complain?
From a software perspective, analog.io is a graphing front end that is backend agnostic. What this means is that users have the freedom to host their own data locally, within the analog.io cloud, or choice of 3rd party services such as data.sparkfun.com. At the moment only phant.io servers are supported but it is entirely possible and on the roadmap to support services such as thingspeak as well.
Ted did have a point in that services like Thingspeak or Xively need to be WAN accessible because they are “in the cloud”.
This isn’t entirely the case with analog.io because it is running local in the browser and can do API calls to a local phant server from that position… I haven’t tested this in awhile, I should try again and write a project log about it!
Good question, analog.io is mostly a Javascript application that is running local in your browser, there is also a database in the could that stores records about you account and references to streams that you have.
However, it is entirely possible to run a phant.io server in a private LAN and only access that data via the LAN with this Javascript app.
Intel also hosts what they are calling Cloud Analytics – provides a place to dump all your time series sensor readings and generate analytic and potentially controls. https://software.intel.com/en-us/intel-iot-platforms-getting-started-cloud-analytics#Setting_up_your_IoT_Analytics_account. Looks like a way to promote use of their devices but appears to be open to any data stream from any device that can talk to their API.
I’m honored to be featured by Hackaday & very interested in everyones comments. Sorry that my original responses didn’t have my actual name in them!
So… you can like access sensors and stuff from the Internets? Been there, done that in 90s with Siemens PLCs so meh.
Awesome comment and super useful contribution to the community. Any possible way you can share your irrelevant work on junky 20 year old PLCs with the rest of us? I’m really interested in turning your brilliant and ancient insights into something useful.
+1 Gerry :) I like the banter, I think this is one sticking point for IoT, the point is not that it is a new technology. Yes all of this has been done before but with advancements with the cloud, internet connections and cheap hardware it is easy enough to use and inexpensive enough that you don’t need to invest 10’s of thousands of dollars in industrial PLCs to make it happen.
Also with smartphones, there is so much more access to this information. 20 years ago, you would have needed to take a desktop PC with a 50lb CRT monitor, hardwired into a physical phone line and dialup to access this info. Now it is just a swipe away on your mobile phone and soon could be right on your wrist.
Love the solar light portion of the build!