Industrial gateway for rural areas with poor cellular coverage

Hello everyone

I apologize for not responding to your comments in time.

The cause is that I had gone on a mission with a team inside the country, I connected using my mobile phone and the connection was very weak, I could not answer but I followed your answers.

And I preferred to wait until I bring you the complete information so that it is clearer for you.

In any case, thank you very much for your collaboration it is the first time that I trust a forum. With the help of your answers, I have understood many things.

I left to do a study of the lakes for which I wanted to set up the LoRaWAN infrastructure (level sensors and LoRaWAN gateways).

The lakes were very large and had depths of 5 to 6 m.

There was GSM coverage from telecommunications operators, mobile communications could be established but the Internet was weak. The lakes were still a few kilometers away from the villages where the offices were to be.

So I saw several types of level sensors on the thethingsnetwork.org page offered by the certified partner companies of the Things Network.

And I thought a lot about the Things Industry Outdoor Gateway as a solution for setting up the first test of this project.

What do you think and do you offer me components that you have already tested (level sensor and gateway) LoRaWAN?

Here are the coordinates of the first lake chosen for a test:

The data changes according to the rainy season:

The area in ha: 4275

Estimated water depth: 2 to 4 m

The width is estimated at 4 km

The length is estimated at 22 km

The distance from the lake to the village is 5 km.

For questions asked by @descartes:

What are the sensor devices going to sense?

The water level in the lake

How many of them?

Depending on the type of sensor or its capacity but I do not know how to define the number of sensors that we must need for such an area? How do I do this?

How often do they need to transmit?

I do not consider frequency a very important factor I think it is depending on what is adaptable to the context, the essential that the information is received.

How quickly do you need the data back at the office?

I don’t mean quickly but according to what is possible and adaptable to our context.

Where is the office? I.e. rural or some distance from the area

The office is in the village where the internet connection is stronger but the lake is a few kilometers from the village.

The information for the moment I want to receive it on the desktop computer using the Things Network LoRaWAN Server and afterwards I must be able to switch to Cayenne.

The question is:

What do you suggest to me for the implementation of a LoRaWAN solution for monitoring the water level and quantity of water in this lake?

thank you
@kersing @jerylcook @pshemek @LoRaTracker @cslorabox

Some of these answers are very vague but you should have some idea about them

Count the number of lakes!

On average, how far away from each village / office are they?

Hourly, daily, weekly, monthly, only when it rains, not at night?

Can you wait a minute, an hour, a half day, a day, a week to get a reading?

What do you do with the water level data - plan ahead, fill the lake up, divert water courses? How you use the data you collect defines context as well.

Can you give the geographical co-ordinates of the lake and the village so we can see what the radio line of sight might be please.

1 Like

Thank you for your reply

There are many lakes and even boreholes, but I wanted to set up for a single lake to see the feasibility of the solution with LoRaWAN.

If I do for a lot of lakes I also have to think about the LoRaWAN network server which must manage all these gateways and sensors and it is also to pay.

Of course I prefer to consult the information in real time and at any time. That is the purpose of this LoRa / LoRaWAN technology. Otherwise it’s not important.

I wanted to use a single sensor for a single lake but what type of this sensor can meet my need?

I suggest you these sensors

That I think effective

What do you think?

Coordinates:

Coordinates of the village: 17 ° 03 ′ 00 ″ north, 13 ° 55 ′ 00 ″ west

GPS coordinates of the lake: 17 ° 6 ‘0 "North (17.10 °) 13 ° 58’ 59" West (-13.98 °)
Altitude lake : 30-30 m

In fact the position of the lake is very close to the village it is at the entrance to the village.

The information is mainly used for planning ahead.

Count the number of lakes!

It’s a lot of lake in almost every village there is a lake

On average, how far away from each village / office are they?

In fact I wanted the monitoring of each lake to be separate from the other lakes.

I mean that for each village I consider a corresponding central lake monitoring office. So for each lake I must have a sensor, a footbridge and a reception desk to give in the village. That’s how I think because the lakes are very far from each other.

I Hope I answered well

thank you

You can make pretty much anything work as a one off project - but there is no point doing a feasibility project for a unit of one without knowing if it can be scaled.

Looking at the area for the co-ordinates and as you have plenty of solar power, it’s flat terrain and only 5km away from the village, it should be feasible with a directional antenna to get data from the lake to the village in this instance.

Once the data is available online, this is entirely feasible as long as you have internet - but it is not normal for a LoRaWAN device to be continuously transmitting.

As I previously lived in a village at 200m above sea level that was at risk of flooding and over the hill was the reservoir that supplied water to Birmingham that some summers you could see the bottom and some summer it was on overflow, I’m aware of the dynamics of large bodies of waters in the UK.

So, how often do you want the water level sensor to send a reading - hourly, daily, weekly, monthly, only when it rains, not at night?

Neither of those sensors are suitable - they are for tanks, not an open body of water!

How many villages then - 10, 100, 1000 - having some sense of scale of the project helps. Rather than finding a solution that is too big or too small - just a rough idea.

Yes, separate lakes being separately monitored - but are they mostly 1 to 5 km’s away from their village or are there many lakes that are 10km away from it’s village?

I’m not trying to sell you anything, I’m trying to un-cover the facts to help you design a solution - don’t read anything in to the questions or try to second guess what is being asked, just try and come up with a numerical or yes/no answer that feels right.

All that said, if you have one lake per village, even with two or three sensors, a LoRaWAN gateway may be overkill. Which is why you need to scope out the project in more detail - no point buying lots of LoRaWAN gateways if other solutions are more practical.

1 Like

every day.

very discouraged to hear this
Can you offer me sensors that can be adapted to my context?

there are many villages that have found lakes.
but we now wanted to study the case for 3 to 4 lakes each in a village.

each lake is a few kilometers from its corresponding village.

Thank you so much

I didn’t understand what you mean isn’t it well written?
Here is the information about villages, each of which has a lake.
these villages are in the same area called Kob but the distance between villages can be tens of kilometers.

lake 1: N 16 ° 33’30 ‘’ W 9 ° 36’50 ‘’
Area ha: 900
lake 2: N 15 ° 86’30 ‘’ W 9 ° 01’70 ‘’
area ha: 1500
lake 3: N 15 ° 50’80 ‘’ W 9 ° 81’70 ‘’
area ha: 4000
lake 4: N 15 ° 74’90 ‘’ W 9 ° 65’90 ‘’
area in ha: 1200
lake 5: N 15 ° 98’70 ‘’ W 9 ° 06’70 ‘’
area ha: 1000
lake 6: N 15 ° 69’70 ’’ W 9 ° 48’30 ’’
area ha: 500
what I think is to put a footbridge per village.
but I will provide information on the exact distance between the village to know if I can reduce the number of footbridges.
what do you think?

these are the Gps coordinates of Kob where there are all these villages and lakes:

Kob: N 15 ° 49 ′ W 9 ° 24 ′

Thank you

Please can you edit your post as it is almost unreadable - I think I can figure out what you are answering - there is the quote option " or you can put the original question in bold, in either case, put the question first, then the answer.

The latitude & longitudes have spaces in them that mean they don’t have to be edited for Google Maps to understand them:

15 ° 49 ′ 02 ″ north, 9 ° 24 ′ 39 ″ west → 15°49′02″N, 9°24′39″W

Not really, water level sensors for rivers & reservoirs are common enough.

I can address the other points once you’ve edited your post - it’s the pencil option at the bottom of your post.

1 Like

If you don’t have decent Internet connectivity (via a mobile provider or whatever) from possible gateway locations “in view” of the sensor locations, then a traditional LoRaWAN network like TTN probably isn’t going to work.

What you could do is get one set of a mobile-backhaul gateway and a few nodes and go try things in some locations.

But from the sound of it, you have the unfortunate combination of a very challenging geographic situation, and limited familiarity with the technology.

Until you have evidence that this idea could actually work, you’ll probably want to keep your investment in LoRaWAN hardware limited.

1 Like

we have the internet connection. only that maybe it is a little weak in certain areas.
Can’t I establish LoRaWAN coverage without needing an internet connection?
Will I be able to make an end-to-end LoRa or LoRaWAN communication and have the information without using the internet or the GSM? what do you think?

really this is what I want to do.
to test that even something and see the feasibility.
but how? what gateway and node can you offer me?

Perfectly.
for tech i read a lot and i think i’m improving but i wanted to try something practical to see what happens.

really you are right.
Thank you very much for your reply.

Not via TTN, and so not within the range of topics covered on this forum.

TTN fundamentally requires good low-latency Internet connectivity from the gateway.

There are gateways which include a miniature stand-alone LoRaWAN network server in the box, generally this is intended as “demo” but it could work on a small isolated network. Chirpstack, RAK, and Dragino all offer this (with Dragino make sure you are looking at actual 8 channel models)

But again those are not on topic here.

what gateway and node can you offer me?

Most flexibility and lowest cost would probably be wiring a LoRaWAN gateway card into a Raspberry Pi; it’s not a very good solution for something to actually leave in the field, but for a brief test it is flexible and gives you some options - which again, are mostly off topic here. You might look at Chirpstack’s Raspberry Pi card image, or try to see if you can actually connect it to TTN via a mobile data plan.

For a node, probably MCCI LMiC on an Adafruit feather. There are many choices, but most bring various challenges so something well supported is key.

To be clear, this is going to be an extremely challenging project, and most of what you need to figure out is not on topic for this forum.

1 Like

I would also add that Dragino also do a low cost open liquid body (vs tank) level sensor - the LDDS75 - I have been testing a couple with reasonable results, and in theory good for a liquid range ‘heave’ of ~30cm to 6m distance so if you lake levels move in that range may be a viable option. A search on the TTN Marketplace will show you others (DecentLabs etc.)

Also if operating remote with poor internet backhaul for the GW you might want to consider contacting Lacuna Space!.. just a thought :wink:

1 Like

well noted.
I will research these topics and especially the gateways that include LoRaWAN servers I have never heard of.
in fact, I have seen a lot of research and IoT project with which I inspired this idea which is:
monitoring six lakes, by studying the position and the number of footbridges that I will place between these lakes, and then configured these gateways on the Things Network’s LoRaWAN network server.
that’s the idea in general.
the problem I have been thinking of is the transmission of data between the gateways and the LoRawAN server of the Things Network using the internet. and if I have no internet and the gateway is placed very far from the office then I wanted to use it LoRa access points to extend the network to the office there will be Internet and the data consultation computer. these access points I have seen proposed by this manufacturer.
Passerelle point d'accès LoRaWAN MultiTech Conduit Access Point - DataPrint.

or to rent with the telecommunications operator special subscriptions or Sim card that I integrate into the gateway even if it will be expensive, so I will have a combination between the LoRaWAN network and the GSM or if there are LoRaWAN gateways with SIM that can be used I will look for.

what do you think this can be possible?

please tell me about LoRa / LoRaWAN simulation software. it can help me to analyze and prove what I think.

Unfortunately
but I will continue my research, it is an objective and it is necessary that I succeed.
of course the help of the members of the forum is much requested.

Thank you

Thank you a lot
it is well noted
I think this is the right solution if I cannot find the gateways include the LoRaWAN server offered by @cslorabox.

This was also offered to me by @kersing but I thought about the cost and the difficulty of setting up.
I will check

Thank you so much @Jeff-UK

@kersing
Thank you it can be a solution adaptable to our context.
how can I have a LoRaWAN satellite for testing? I hope it is not very expensive?

What is your budget limit for the POC - thay may influence recommendations… :wink:

1 Like

the budget is according to the efficiency of a solution however it should not be expensive as this makes the solution inappropriate.

Thank you a lot
your answers have helped me a lot.
@Jeff-UK @cslorabox

You keep giving non-numeric answers that don’t help us help you.

So, is $250 too much, about right or, as that’s a pretty low number, you could run to $1,000?

What is the value to the village to know the lake level? How will the costs of each installation (hardware & time) be recovered?

1 Like

I am asking for your help but the problem is not the cost now. the problem is the feasibility of the project.
in fact me i thought to find a solution, to do its test for two or three lakes and then to scale it (here other things must come into play).

I thought that my analysis after the research I did that my idea should be feasible but because of your comments and especially that of @cslorabox who advised me not to invest in LoRaWAN equipment as long as I Didn’t find a solution to prove it, made me go back, do it slowly, and do some little testing or prototyping with the low cost components.

this is why I thought of ordering the dragino Lg308 and some sensors to see what will give and to be able to become more familiar with the LoRa / LoRaWAN technology.
@cslorabox the Chirpstack has good documentation but I haven’t seen its components.

the information will be exploited by those responsible for this sector.

Thank you @descartes

There are gateways you could run some tests with at $99 or you could buy a gateway for $1,800 and at numbers in between. The point of the question is to guide you towards buying something appropriate.

The ONLY definitive way to prove that a LoRa based solution is not a solution is to try it - so you will have to ignore the advice to not invest in the technology - if you don’t, you won’t be able to try it out.

Personally I think there are so many use cases for a LoRaWAN network to cover a village & surrounding area - monitoring livestock feed & water (or even location), soil moisture levels and so much more, that this is a bit of a no-brainer if it were fully fleshed out.

But if it is only going to be used monitor water level, then LoRaWAN would be money spent on tech that would be far more capable than the simpler LoRa P2P link that would be needed.

You understood well.
I am looking for the components to do a test with an affordable cost.

The objective of this topic was to offer me a gateway that is the most adaptable to the rural area according to your experiences and which works with the Things Network LoRaWAN Server and all other complementary components because I wanted to test the LoRa technology myself and then decide the implementation in place of this project at scale with the Things Network server LoRaWAN.

I fully agree with you.
this is what I really wanted to do, even if the solution does not achieve our objectives we can use the components for other use cases we lose nothing.

of course, if we set up an efficient LoRaWAN network we can use it for several areas:
for example,
-Know the real-time use of water, for example, if livestock uses this water, what is its consumption?

  • can be used for agriculture
    -for the surveillance of the position of the cattle
    -maybe even to monitor the health of livestock.
    infinitely

what I think
if I am going to do a test for a single lake the LoRa P2P must be sufficient.
but if I want to do a test for six lakes away from each other I have to use the LoRaWAN and determine the location of the best suitable gateways to cover all six lakes.

What do you think?
And your proposal for the gateway?

Thank you very much for your collaboration @descartes
Thank you for all

I will happily make some recommendations but I’m not reformatting the lake co-ordinates for you:

Once you have done that, I’ll look at distances and sight lines for you.

Deal?

1 Like