Topic: My dream weather station

[SLOweather]

  I don't know if it is a myth or urban legend but I've heard of a chocolate bar melting in a guys pocket but maybe he was just overheated otherwise.

From  https://www.technologyreview.com/s/400335/melted-chocolate-to-microwave/:

Raytheon credits the discovery of microwave cooking to a grade-school-educated engineer named Percy L. Spencer. One day in 1945, Spencer was walking through a radar test room with a chocolate bar in his pocket; he came too close to a running magnetron tube and the candy began to melt.

Spencer was already known at the company for his curiosity and ingenuity, and he reacted true to form-he pointed the magnetron at kernels of corn and watched them pop, aimed for an egg and saw it cook so quickly it exploded. Soon, Raytheon officers were sampling microwaved meals in the executive dining room and within a few years the company was marketing the Radarange for $2,000 to $3,000.

Hmm, 40 years ago I unknowingly repeated the egg experiment with an Amana Radarange in college. Just as we were trying to figure out how the tell if the egg was cooked, it went off like an M-80...

[SLOweather]

This is something I always found funny... I personally don't have a smartphone, but if I had one, I can guarantee you that I would much more prefer a thicker device, but with a decent battery and connectors that I have accessories for.

Heh... The first thing I did after I got my Galaxy S4 was to order an extended battery and back. It lasts longer, fits my pocket better, and is easier to hold. I've gotten extended batteries for just abut every phone I've ever had. I even had a AA pack for one of my first brick phones.

[Josiah]

Regarding the original topic:

If you wanted the sensors to plug into pysical "Hub", I would think that you would need a small micocontroller on each sensor. The microcontroller would handle reading the sensor directly and provide the read value to the "Hub". The microcontroller attached to each sensor would also contain some basic "properties" so that the "Hub" could know what type of sensor it was dealing with (eg. temp, wind, precip etc.). The "hub" would in turn provide the data to wherever it needed to go.

If someone were to create something like this, the first task would be creating the "Sensor Microcontrollers" for existing senors available on the Davis, Peet Bros., Ambient (etc.) weather stations.

It's not a far-fetched idea and is definitely possible given the technology available today.

[Jáchym]

Yes.... all the technology is available, it is just a question of putting it all together.... Just like the iPhone for example It was a revolution, even though everything it had already existed - touchscreen display, MP3 player, a phone, a digital calendar... Apple never invented a smartphone, they "assembled/built" it

[Josiah]

I don't doubt what you are saying. I am thinking along the lines of actual implementation.
Another thought... Make a source code template for the "Sensor Microcontrollers" open source so it's easier to create custom sensors for those who want to do so.

Maybe have each "sensor microcontroller" programmable, use a "stackable" communications interface to the "Hub" (ie you don't need more connectors on the hub as you add more of these "sensor microcontrollers") think something like I2C but without hardcoded IDs, this would allow you to connect any combination of (and multiples of each) sensor.

Or maybe use I2C but with external DIP switches to allow maunally setting the cummunications ID..?

I wonder how much interest there really is in something like this...

[Bushman]

We need a weather version of NMEA 2000 - but wireless.

[WeatherHost]

Self powering using solar and wind with power cells that never need servicing.
Self cleaning and maintaining.
Accurate
Temp, Humidity, rain, wind, snow, fog ..... all monitored from three or more remote stations and averaged.
Full color display console.
Fully configurable PC reporting and graphing
20 year unconditional full replacement warranty.

Under $50

[Jáchym]

Self powering using solar and wind with power cells that never need servicing.
Self cleaning and maintaining.
Accurate
Temp, Humidity, rain, wind, snow, fog ..... all monitored from three or more remote stations and averaged.
Full color display console.
Fully configurable PC reporting and graphing
20 year unconditional full replacement warranty.

Under $50

Nice Though Im afraid this really is a "dream" station

[WheatonRon]

So, yesterday I couldn´t fall asleep and I was wondering what would my dream station look like. And I came up with a few ideas…

So, the major concept would be modularity – i.e. there would not be a fixed set of sensors, console etc. The whole station would primarily consist of some sort of a unifying hub, to which various devices could be connected.
Let´s start with the sensors. There would be a whole range of different sensors and for some parameters even more types. So, for example there would be several rain gauges, from cheap plastic ones, to high-end expensive weight rain gauges. Likewise, there would be several types of temperature sensors, starting with some cheaper, plastic and not so well insulated one, to some very good durable and actively cooled sensor.
In other words, you would be able to buy those sensors that you are interested in and even later on change them, so if for example you could not afford the most expensive one, you could buy the cheaper one and then later on change (or add) just the more expensive one, rather than having to buy an entirely new station.
There would be many different sensors and the biggest advantage would be that they would all be connected to that one single hub, so the output would contain all these values as one single dataset, which could then further be processed in various ways.
I thought about the following sensors:
•   temperature – air, soil
•   humidity – air, soil moisture
•   pressure
•   wind speed/gust
•   wind direction
•   precipitation (optionally heated)
•   snow depth
•   lightning strikes
•   emissions – CO2, NO2, O3, PM10 etc.
•   solar radiation
•   UV
•   sunshine duration
•   seismogram
•   evapotranspiration
•   webcam
Now as I said, you could basically combine them in any combination, or even have several at various locations. The great thing about it is the fact it would all be one dataset, even the webcam image could then be easily paired with the rest of the data.
For all those sensors it would be possible to either have them connected by a cable or wirelessly, and likewise, powered by either a battery or solar-powered.
Now that is the first part – the input. And now for the output.
That would more or less be the same concept. There would not be any console, you would be able to attach various modules to that hub, that could send the information using various communication channels.
So, for example there could be obviously a wi-fi module, which would immediately stream the data to your own server, to your NAS, to cloud services, weather networks etc. Then there could be for example a Bluetooth transmitter, which could stream it around your house and I can imagine there could for example be an app for your smartphone/tablet/monitor/TV and you could simply have a live view of your data, just like a console – put a cheap tablet somewhere in your house.
It would also be possible to attach a module for a memory card to the hub. So all data would be saved to the memory card and probably there could be two cards. One acting as a backup, so for example under normal conditions it would be saved to just one, but in case your internet connection was down, it would automatically start saving data to the other one too, so that you always have the data saved to at least two locations. Or you could use that to for example transfer data to your PC if it was not connected to it 24/7. You would simply take out the card and while you would be transferring it, it would be saving data to the other one and once the other one was back in place, it would copy whatever it recorded while the other card was disconnected.
Then there would be modules that could communicate via the MODBUS port, so you could connect it to various other devices in your house, create conditions and activate things like irrigation, lighting, heating,  etc. This would include some sort of a universal API or XML protocol. The hub would be something similar to computers such as Raspberry Pi, so it could create these.
I know it is a lot, but remember – this is just my “dream”, not something I expect to see in the near future But the best thing is you would be able to gradually invest more and more and get more and more features and have it all integrated and interconnected. Why invest in an expensive console when everyone already has a TV, monitor or a tablet at home. Why buying a completely new station just to get a better sensor that you could not afford or was not available before. Why having separately detectors, webcams and sensors all sending data at different times, via different channels etc.

I just found (September 29, 2016) this thread and realize it is several months old but very interesting. One of my frustrations with this Forum is finding material as the search function does not appear to be as "smart" as Google's. In any event, good post! Take your ideas and those suggested by others in this thread and send to Davis! Maybe the making of the long awaited VP3!!

[Josiah]

I've been toying with this idea (along with WxTech's Dream Machine) for awhile. I think there is a way to actually do this.

Apparently in 2015 the Open Geospatial Consortium (OGC) published the SensorThings API.
Basically it's a communication standard for multiple numerous "Things", each with one or more sensors, uploading data to a system. In addition to the uploading side for the "Things", the standard also specifies an API for later retrieval of the data.

Quote from Wikipedia (for context as you are reading: it helps to think of the "things" as the devices reading from sensors; which could be existing weather stations):

SensorThings API is designed specifically for resource-constrained IoT devices and the Web developer community. It follows REST principles, the JSON encoding, and the OASIS OData protocol and URL conventions. Also, it has an MQTT extension allowing users/devices to publish and subscribe updates from devices, and can use CoAP in addition to HTTP.

The foundation of the SensorThings API is its data model that is based on the ISO 19156 (ISO/OGC Observations and Measurements), that defines a conceptual model for observations, and for features involved in sampling when making observations. In the context of the SensorThings, the features are modelled as Things, Sensors (i.e., Procedures in O&M), and Feature of Interests. As a result, the SensorThings API provides an interoperable Observation-focus view, that is particularly useful to reconcile the differences between heterogeneous sensing systems (e.g., in-situ sensors and remote sensors).

An IoT device or system is modelled as a Thing. A Thing has an arbitrary number of Locations (including 0 Locations) and an arbitrary number of Datastreams (including 0 Datastreams). Each Datastream observes one ObservedProperty with one Sensor and has many Observations collected by the Sensor. Each Observation observes one particular FeatureOfInterest. The O&M based model allows SensorThings to accommodate heterogeneous IoT devices and the data collected by the devices.

Unfortunately I have not been able to find very many implementations of the API, but here is a demo dashboard by SensorUp which focuses on the back-end "wiring" side.

What I'm getting at is something like this would allow someone to start with an existing weather station pushing data to the system and later add (or remove) other sensors of whatever types and in whatever combinations they choose.

What do you guys think?

[vreihen]

I know of a handful of people using NodeMCU Arduino boards to run IoT sensors, although maybe not following standards.  I built a UV/sunlight sensor, a barometer sensor, and am working on a rain gauge right now.  All are augmenting my existing station in weewx.  For under $8.00 with 2.4Ghz wifi built in, you would be hard pressed to beat the NodeMCU.....

[gdreelin]

I agree 100% I have had a WH1080 station for years (though not accurate) I have wanted to build from the ground up. I have good dedicated Weather Display Rackmounted box that backs up to to a very large NAS. Now just want to get better sensors but there are so many choices and not sure where to go. Sparkfun has nice stuff, Adfruit, Photon, Pi, etc. The issue I see is comm back to WD from different types of sensors. 1-Wire made it some what easy but over the years I dropped out of weather it looks like that has gone away to some degree. I am all for making my own stuff then I know it is something I can rely on. The idea of going to a tablet is good working with many IoT devices that do that now love to IoT the heck out of weather stuff.   

Sent from my XT1650 using Tapatalk

[gdreelin]

Yes it is absurd, because everything I said already exists, there is nothing that would need to be "discovered".

The problem is the integration. All the above exist separately and the best thing about my dream solution is that you work with one source, one dataset. You dont need any console, Im sure many of you for example have some old tablet or smartphone, which is more less useless, but has a fully functional display, so why not re-use it. Or there are devices that can measure the emissions and even send them to your server, but.... again, first it is not synchronized with the rest of your measurements, but second, it is quite expensive, because you are paying for another wifi transmitter, another chip...

On the other hand maybe this is exactly why no-one has developed such a universal device. Because you could then choose if you want thermometer from manufacturer A, rain gauge from B and use an LCD monitor from manufacturer C to display the data.... obviously they will always try to make proprietary solutions, where you have to buy the whole package from just one company.

OT: this is slightly off topic and take it just as a personal note, but I am a big fan of Google, use it all the time for my calendar, email, documents etc. And meanwhile I really dont like Apple. And I am not one of those "haters" who never even saw the device, just hate it for the sake of it. I actually "had to" use it while at the uni and the more I worked with it the more I hated it. But ok, I fully respect that others might think otherwise. But one thing that I really hate about Apple is how they try to make everything so that you can only use their over-priced products. There are rumors that the next iPhone wont even have the normal 3.5mm audio jack - Im sure everyone will love them for this given many people have expensive headphones and now will either have to buy expensive ones from Apple or use a reduction which will degrade the sound quality.

On the other hand, it is understandable... this is how Apple makes sure it retains its customers, especially given the fact that since Job's death, they didnt really present anything revolutionary and their sales are going down. It is in part also because of their closed platform.

Check OpenHAB it helps integrate sensors and free!!!  

Sent from my XT1650 using Tapatalk

[gdreelin]

There are a lot of groups now integrating sensors together since many got tired of buying Philips Hue, WeMo, ECO, and other IoT devices that could not communicate with each other. A standard came out for things like   Amazon Echo, Google Home, and or Smartthings that allowed them to control them. As always people wanted open source so groups like Hackster.io, MySensors.org, and OpenHAB have created controllers to combine all these items. Been running OpenHAB for year and works great so far.

Sent from my XT1650 using Tapatalk

[Bashy]

2 add on sensors that are important to me, so much so that when i upgraded to Davis
i had to buy the Soil, leaf, temp station and the leaf wetness and extra temp sensors
to be able to carry them over from my OS station and that is the Night Cloud and
Rain Duration sensor, the latter is really more about setting the website current icon
is quickly as possible and not having to wait til the 1st tip and then not have to wait
til X amount of minutes to change back once the rain has stopped, this made even more
responsive by adding a heater to dry out the leaf board. As for the night cloud, speaks
for itself really, i wanted something where the website current icon would still change
during the night Had i not been able to achieve at the very least, the RD unit with a
Davis station, i would never have spent the Ł2k+ (all in) on the Davis setup.

I also think a good station should not be susceptible to interference either, and many
stations suffer with it, what with sensor drop outs or the whole thing flatline, no names,
DAVIS....... although i cured my flatlining issue but there are many that still suffer with it.

[Bushman]

Yes.... all the technology is available, it is just a question of putting it all together.... Just like the iPhone for example It was a revolution, even though everything it had already existed - touchscreen display, MP3 player, a phone, a digital calendar... Apple never invented a smartphone, they "assembled/built" it

Actually, the IBM "Simon" phone released in 1992 was the first touchscreen.  Apple merely copied the idea (like most of its other products) and made it "better" (subjective)

[Jáchym]

Thats exactly what I said Apple always just gathered ideas, put it all into a nicely looking product and added perfect marketing and "corporate design".