Coming from a Freifunk background I am wondering whether the purely technical scope of TTN means that we miss a great opportunity to not only collect data but making it available to the public.
Some of the data collected on the community network may be of interest to others. In other contexts we are discussing with the governments to release publicly acquired data into the public (Open Data, Open Government). So why not make the Open Data principle a first class principle in TTN? Think Citizen Science.
This does in no way imply that information collected via TTN must be public.
I understand that TTN is neutral to all uses - also to private ones. This is fine and security and privacy are very important.
But where this is not the case I think making available all those sensors (and actors, but this is much more complex) to the public can be the foundation of a lot interesting follow-up work (based on aggregation and analysis).
Just an example: The few beekeepers I know are all confessed tinkerers. What if you not only have data from your hives but from 100s? Weather and pollution are also low hanging fruits.
TTN is also a great match for other projects like SenseBox which has a data aggregation backend (OpenSenseMap). Iāve started the conversation with the SenseBox people.
Hereās my proposal:
Letās have a discussion
Consider adding this to the mission/manifest (if people agree that this is a good idea).
Build infrastructure for it
IMHO the latter should start with a Data Catalog. Which may start as a wiki and evolve into something that checks if feeds are live, categories etc. I would expect that people converge on data formats eventually.
Thatās exactly what the corporate ācloudā overlords like , with your data, because big BIG data means profit for some and overregulation and loss of privacy for most of us.
@BoRRoZ just observing it looks like this is the way society is heading for anyway. It would be very ironic if useful applications of open data would get stuck on this limit-big-data-where-possible idea, while it would not limit in any way corporate initiatives that are a bit less useful to the general public.
There are many possible applications of open data that would not harm privacy. I think brainstorming about applications wouldnāt harm. Besides it looks to me that currently solving the technical issues is far ahead of developing useful applications, and this means that it might be a good moment to extend the scope.
Letās start with the first question in your title: āIs TTNās scope too small?ā
I think the scope of TTN is actually quite big. Building a ātechnology independent IoT networkā means adding other radio and IoT technologies next to LoRaWAN, which will already be quite some work.
I think itās good that TTN limits itself to being a network. You could see TTN as a set of WiFi routers, connecting objects to the internet. I donāt think itās the job of the router to intercept all traffic for the sake of open data. Applications on the other hand are perfect for this. We already see great open data initiatives such as ttnmapper, a community project for measuring the coverage of TTN, or safecast, a crowdsourced platform for measuring radiation levels (and has prototypes of their equipment running on TTN).
So āDo we need to go beyond pure technology?ā
Absolutely. But I suggest to implement it on top of TTNās routing services. Just like your WiFi router should focus on routing packets, TTN should do that as well. Aggregating and visualizing that data can be extremely valuable (and I think itās super interesting), but itās the responsibility of a higher layer application or even a completely different platform. Hereās how I think this would look:
I like the idea of a wiki page that lists open data initiatives. Feel free to start collecting those and create a page on the wiki! Common data formats is absolutely a good idea. Sensors usually send a few bytes of binary data, that can be decoded and converted to JSON using the āpayload functionsā (see the dashboard).
But if someone wants to know what that packet with payload 'F3 means, its the door from my 3e freezer in the cellar, every time it opens a signal is transmitted through the LoRa gateway, telling the backend to start a timer.
Thatās a very interesting point of view and I never considered it this way. But it also makes me wonder what exactly the definition of TTN would be. Thereās a core development team which developed the network as it is at the moment. And this is a massive achievement and I can imagine there are still a lot of possible features to add. So according to that I fully agree that the scope is by far large enough.
But this core team is only a part of the TTN community as a whole. In fact most forum members do not develop any infrastructure-like things and they also will never do so. Because of their skills, background, interests, and so on. Wouldnāt this allow for space to extend the scope anyway? It seems that the interaction and thus cannibalism is very, very limited. I donāt say that it would necessarily be something in the open data area. One could argue that anything that contributes to a network roll out would implicitly fit in the scope.
One could also argue that TTN provides (free or not) the network and infrastructure (together with gateway owners), there are rules howto communicate with and over this network.
The moment TTN forces a user to fill in a form (as a rule) which explain exactly what his/her payload means, catagorize this payload and label it with a personās ID, that moment it will loose its current and coming attraction imho.
Its like the watercompany forcing you to explain what you did with that last 6 liters of water.
And in fact that is what youāll have to do if you want to sell/use this āprivateā generated data in a usefull way, but off course I can imagine use case like air/water quality for example, were this information can be important ā¦ to some.
In the near future I even canāt post this message without using my governement supplied internet ID card, how this will work out for āopenā networks like TTN ? ā¦
nobody said anything about āforcing users to do disclose anythingā.
At least this was no my intention.
Itās about enabling the community to use the data that we - as community - collect to be something more than our personal benefit/delight/experiments.
If itās pure transport then there is a bit of motivation -besides the ābecause we canā - missing or it feels incomplete for me.
People spend time&money into building this. To which benefit? Is there a non-technical goal? I think there should be one.
I agree with @htdvisser that this could be a different project that runs on top of TTN as well. I plan to integrate on of the SenseBoxes with TTN and will try to develop the idea a bit further with the local guys from the OpenKnowledge Foundation.
me too ! and donāt get me wrong, Iām a big big fan of this fantastic project !, but Iām personally rather 'sensitive about the direction the internet as a whole is headingā¦ (and Iām there from āthe startā )
I agree as well. I think it makes sense to have the limited resources in TTN focus on building and providing the network(s) to move IoT data around. Another project can use TTN and other networks to process data and make it available.
Have a look at the Manifesto and Mission of TTN. As youāll see the definition is also pretty much included in the name: The ThingsNetwork. We are a Network that connects Things.
I think it could be really awesome to start The Open Data Organization on top of The Things Network. This way, TTN can focus on the network part and TODO on the open data part.
You would probably choose a better name than TODO thoughā¦
I did not have an full blown IoT service stack in mind, but just the ability to access data from the network. Iāve formulated a feature request to this effect:
The Things Network is not designed as an IoT cloud platform, but as a developer friendly network services to build devices and applications.
Iām currently working on the design of the integration architecture. It entails traffic in both directions including metadata, device registry synchronization and status monitoring. Weāre building reference integrations for Azure IoT Hub, IBM Bluemix, AWS IoT, OpenSensors.io, etc. Youāll be able to deploy the integrations from The Things Network Dashboard.
The platform that youāre looking for is probably OpenSensors.io. Iām working with @yods on that.
BTW, the community is free to build something on top of The Things Network, connecting directly to Handlers and grabbing, storing and publishing data. But this problem has been solved already, thatās my main point, so weād like to integrate first.
I agree that TTN should not be an IoT platform and thus IoT platformish features should be avoided. And perhaps #216 went too far in that direction.
But thatās not what I asked for (in #215, which has been closed now)
I was looking for simple open message access - a lack of authentication.
This is hardly an IoT platform feature.
I disagree with you dismissing this offhand.
Building integrations with commercial IoT platforms will be very valuable.
But I donāt feel this to be the right approach since e.g. this will cost money and/or force their licenses down your throat (in case of opensensors.io).
