Sparkfun enters the IoT mix
Buck Rogers
Posts: 2,185
Hello!
Last week Sparkfun announced that they were creating a new site, which would enable people to do stuff with their data. The site in question is here:
https://data.sparkfun.com/
Of course on the masthead for the site they list what sort of systems they expect to be connected to the site. And they describe the code it runs and that its also portable. My original idea was to make use of a Quick Start board, and this guy, http://www.parallax.com/product/40002 that's the "Wiznet Ethernet Board for QuickStart". While I'm not sure about composing the code the pair will be running, I just find it interesting that Sparkfun left out the Prop family of boards for the whole mix.
Last week Sparkfun announced that they were creating a new site, which would enable people to do stuff with their data. The site in question is here:
https://data.sparkfun.com/
Of course on the masthead for the site they list what sort of systems they expect to be connected to the site. And they describe the code it runs and that its also portable. My original idea was to make use of a Quick Start board, and this guy, http://www.parallax.com/product/40002 that's the "Wiznet Ethernet Board for QuickStart". While I'm not sure about composing the code the pair will be running, I just find it interesting that Sparkfun left out the Prop family of boards for the whole mix.
Comments
Show up on their new site with some really cool IoT projects and people will want to duplicate them. Easiest way to duplicate them is to use the same parts you did and run your code. If folks are more ambitious, then they can translate hardware and port software.
Get out there and do things!! :0)
It's like going to your boss - show up with a solution, not just another complaint or problem.
So OBC's sig line means "announce projects to the world beyond the propeller forum, but not platforms" ... otherwise it is rather hypocritical.
I really don't like getting to the end of the day regretting I had innocently posted something to have it turned into a dig against another member when that was certainly not my intent in any way.
C.W.
I like it.
On the idea itself, as a researcher or experimenter, I wouldn't want to use someone else's data without more knowledge about how it was collected, and the instruments used. Who knows if it's accurate? Most people's data is flawed one way or another, and it takes considerable work and effort to ensure its validity. This seems rather like Wikipedia with absolutely no peer review for anything posted.
While I think the concept has merit, it's the *collection* and analysis of data that is of scientific merit. I can only see the raft of meaningless applications based on questionable data. None of the examples I clicked through had any comments whatsoever on how the data was sampled, filtered, or even if was reviewed. One I looked at, for a DIY Geiger counter, didn't mention if the tube was covered to not accept alpha particles. So, not very useful, as we have no idea what the tube was receiving.
Common sense and informed gut feelings are undervalued these days IMHO.
"Internet of Things", you know, I'm not sure I want my things on the internet. Like my children I would never know what my things get up to on the net.
Trust me, you really don't want to know. But if it ever gets internet access it's going to take over the world.
Or maybe random data has value after all.
SparkFun is making this very easy. And free. Which is a pretty good idea if you sell all the required parts and want to sell more of them!
Is this data any use to anyone except the guy generating it? No. Although it may be useful if aggregated and used as entropy for /dev/random
By the way. The light in my fridge is now off. In case anyone is wondering. Of course I can't really tell for sure unless I get that goo in my fridge a twitter account.
Then it's pointless to make it public and make a big deal of listing the data streams. It's obviously being done this way to generate a buzz. Well and fine, but if the data is to be public, at least make the effort to tell others what the data actually is. Each data set already has a short comment. Would it really have been asking too much for a few more fields on how the data was collected, so that others could determine if it's better than just generating numbers at random?
I'm jaded in that this seems little more than an Internet version of the publicity stunt. There's nothing wrong with that, but in the absence of any practical use of the data they are taking steps to publicize, call this for what it really is.
Take for instance, there is a device called a PCR that is an open source kit for sampling DNA and sells for $599. Now a guy has posted a project on the Hackaday Contest costing ~$5 although now he knows he needs to add a processor chip.
http://hackaday.com/2014/07/14/thp-entry-polymerase-chain-reaction-cheaper-than-a-hamburger/
Point is, if this guy succeeds in making a really cheap DNA sampler for even $100 it would be fantastic. At least this is going to inspire peoples imagination to solve the problems described.
If successful,it could then be built by schools, and who cares quite how accurate it is as long as it can inspire the school kids. Of course, any hobbyists interested in this field could also have a ball experimenting. Who knows what this will result in??? I am watching it closely as even I am interested.
My point is, anything posted on the internet has the capacity to change the way we think about a problem, and may just be the missing link to a simpler or cheaper solution that was ever thought possible. The more of these types of places/forums/what-have-you on the internet, the better off we will all be.
The medical advancements have accelerated due to the internet. I have a (single) personal experience/proof of this.
The old satellite that they tried unsuccessfully to recover was inspired by the internet. Communication was rebuilt and re-established, but sadly the engines failed to fire.
The "Projects, not Platforms" quote is probably getting a little dated, but I think it still holds as a good focus reminder...
At one point it seemed that we had a couple dozen different options, but nothing was making any traction outside of the Propeller Forums.
The cost of Quickstart forced my hand into platform design, or I'd probably still be generating stackable boards...
Once we get design #2 out to print next week, I'll be heading back into a more project direction myself. (it's more fun.)
Now about the unhappy bird, can you post us a link?
And since I started this off, what my proposal happens to be is one of running the service locally, on a RasPi as it happens, and feeding it data from a QS wearing its gizmo containing an Ethernet device, and a webserver. Only problem here is that I don't know if the combo supports outside data connected to the I/O points on the Prop board.
Incidentally I am naming a random individual to the lists of problems that only Murphy would love.......
That is to say if communication only goes from your sensors to your Propeller to your Raspberry Pi then how are you going to check the status of that goo in your fridge at home when you are away on holiday?
Generally people cannot access services running on machines at home from outside. They don't have a fixed IP address, they don't have a host name, their ISP blocks all the ports, they don't have the necessary know how.
Hence the need for all these "cloud" services.
I actually think this is a shameful situation. The "internet" is supposed to be a "peer to peer" system. If ISPs don't allow my servers at home to be seen from the outside then that is not the "internet" they are selling me.
Yep, that is what goes on here. Luckily ISP's in this part of the world don't seem to worry much about running service on you home machines. In fact they are I kind enough to provide 10Mbps up and down speeds for all of 5 euro a month!
Still, most of the worlds population are not into that level of net sophistication. Hence things like the SparkFun effort.
Nice to know you can get ten by five for five euros. In my line of work it helps to have to maintain that perspective.
Still not sure what to think of SaS like SparkFun, give me a couple years and I'll have an opinion...
Seems to me that is not what SparkFun is offering.
They have some server software that is open source and usable by anyone on any machines they like. It's a simple node.js app https://github.com/sparkfun/phant
They are running this on some servers somewhere and allowing us to use it for free.
At this stage the whole thing is barely more than a quick hobby project for fun. Although, as I said earlier, such "fun" will stimulate sales of the various WIFI, ethernet, whatever dongles for your "things" by SparkFun.
In the Open Source Hardware and software world that SparkFun operates in they would not get very far offering SaS.
Speaking of Open Source Hardware. This modern enthusiasm for OSH is kind of amusing.
On the one hand it is no new idea. Since the dawn of electronics there have been magazines where readers could submit articles about their designs and creations. It was always expected that anyone could build them, improve on them, for themselves no matter if it's for hobby or commercial use. All that happened is that the dead tree magazines got replaced by the internet. It only seems like a radical new idea to the new generation that grew up in the dark period of the PC and proprietary everything, when for various reasons electronics as a hobby was almost extinct.
On the other hand, there is almost no such thing as "Open Source Hardware". At the heart of all those OSH gizmos from the Arduino to all the little boards sold by SparkFun, AdaFruit and co are very closed source micro-controllers and other micro-chips.
http://arstechnica.com/science/2014/07/recently-revived-isee-3-probe-cant-fire-engines-cant-change-orbit/
Sure, data clouds work around the limitations of some ISPs, and unless you're one who believes everything on the Internet should be free, these services have been around for years. I applaud SF for offering it for free, but the unique situation they're in -- heralders of open source, yada yada -- calls for a more pertinent expression of those ideas. People don't have to do a full blog write up, but if you're going to defend such a service as offering inspiration, let's see the concepts and methods behind this data. Take TWO MINUTES to fill out a little form describing your project before they let you create the push channel. Is that so hard?
If anything, their data service will inspire someone to do it properly. As SF has it, this will be used by a core group, and ignored by everyone else as irrelevant. There is no inspiration from just looking at the nth listing of some dude's local temperatures.
Perhaps you are over thinking all of this. Imagine this scenario:
10 year old boy or girl has just worked through all the suggestions that come with the SparkFun Inventors Kit that they were bought for Christmas.https://www.sparkfun.com/products/12001 Being keen and imaginative they have implemented some brilliant gadget, say a cat flap detector. They can tell when their cat goes in and out of the house.
Now they get their parents to buy some WIFI dongle for their creation, or ethernet. And now they can push the movements of their cat to their own log on the net, conveniently maintained by SparkFun. Easy to use and free. Brilliant!
If I was that 10 year old I would be well pleased with myself.
I can't be totally sure but my guess this is the kind of target audience SparkFun has in mind. As I said it stimulates sales of their hardware dodads. It's all good.
As you rightly say, any serious distributed data collection and measurement would require far batter organization and traceability.
God I wish I had that inventors kit when I was 10.
Apart from "who cares," this still doesn't answer why make such data streams public. The kids and the cats don't care. It's public only to serve SparkFun -- makes a buzz. Okay, that's fine, but don't call it inspiring. Five minutes on Hackaday is inspiring.
I guess I'm not explaining this well. This isn't about data clouds, free or otherwise I'm disappointed Sparkfun took the easy road. There's so much more that could have been done, given just a little more work to turn this into a solid educational experience.
The kids might care, they can show off their creations to their peers over the net without the hassle of signing up or logging in etc. Simplicity has great value here.
I don't know. It is what it is. People might play with it for a bit just for a bit of fun. They will have learned something simply by getting their Arduinos or whatever on the net via some WIFI dongle. Perhaps that's it. Perhaps it inspires them to pursue other approaches to networking stuff. That suggests there is an opportunity for Parallax to explore here