Hello , I want to implement LORA netwotk for measuring Silo level


I’m interested in Ultrasonic silo level measurement I need it to sense pekan nuts level on multiple silos.

I have no previous knowledge on how LoRa networks work, I did some research but it’s not clear to me what else do I need to make an end device sensor work as an IoT device please note I don’t have any infrastructure such as gateways/nodes could and I’m not clear on what I need to get started testing.

I attached pictures of my client’s plant, please note there is 200 silos and its a

electrically noisy environment
I’ve researched and liked these sensors from [https://www.tindie.com/products/edwin/llds12-lorawan-lidar-tof-distance-sensor/]

So, what else should I need to get started testing besides the sensor ?
How can i collect data?
I appreciate any help :smiling_face_with_three_hearts:
I attached some pictures

WhatsApp Image 2020-08-10 at 4.26.05 PM (1) (1) WhatsApp Image 2020-08-10 at 4.26.05 PM (2) WhatsApp Image 2020-08-10 at 4.34.26 PM (1) pekan nut

I’ve done a lot of work in Grain Silos. The first question, does your client require the electrical equipment to be Dust Ignition Proof or something similar, for example Intrinsically Safe?

Looking at the light fitting in the second photo, I suspect there could be some requirement.

This will become the first filter you use when selecting sensor hardware.

1 Like

Do you already have any kind of level indicator? If you are lucky you could get data from an existing sensor.

If you don’t already have a sensor, what kind of data do you want to collect? e.g. silo level (analog from empty to full), silo weight (0 kg to big-number kg), silo level limit switch (on/off = full/not full).

To be perfectly honest, because all of your silos are in a factory and there is power and cable ducting present, I would probably not make LoRaWAN my first choice. In this environment I would look at some kind of industrial plc system and a central controller. You have gantries and catwalks to access the silos and to run your wires. A simple RS485 multidrop network (for example) would let you collect data from several silos with one run of twisted-pair cable. For 200 silos you probably need several RS485 links, but you can muliplex them, maybe 8 links with 25 silos per link. Furthermore, I’m guessing that this factory has an internet connection in the control room, so you could easily export the collected data to somewhere else.

It’s very tempting to want to use the cool new stuff, and LoRaWAN is pretty cool. However, don’t forget the tried and tested techniques that have been used in the past. They still work.

The equipment should be lets say IP67 since its outdoors even though it has a ceiling dust wind and rain could get in

yes analog form empty to full so i can display the silo level remotely i don’t need very high accuracy

And yes they have electrical and internet in the plant but there its lots of costs involved using the traditional methods of PLC

The initial trail testing should be of 10 silos and we will see from there.

I think you’ll find there are lots of costs involved with LoRaWAN.

In general, LoRaWAN is ideal for devices that are remote, or hard to get communications wiring to, and which send a small amount of data periodically. Again, I suggest that your use case would be better solved with something other that LoRaWAN.

Let’s look at what’s required:

Sensor (e.g. ultrasonic distance sensor)
LoRaWAN node
LoRaWAN gateway

If you are lucky, you can buy an off-the-shelf sensor+node. You need to provide power to each node. They can be battery operated, but that is usually for nodes that send data infrequently. If you send data too often you will run the batteries down and need to replace them often. If you have 200 nodes that will get very annoying very quickly. So, probably the nodes should be mains powered.

The gateway needs to be located nearby, where the nodes can reach it. You don’t want to rely on someone else’s gateway, so install your own. Probably in the control room where there is power and internet. The gateway needs a network connection.

You don’t have to use an off-the-shelf plc. I agree they are expensive, but an Arduino with an RS485 transceiver might work for each silo, and a Raspberry Pi might work for the controller. However, if you do things too cheaply then it’ll cost more in maintenance. If this factory has 200 silos which are full of pecan nuts from time to time then it seems there is a lot of money floating around.

‘Conventional’ data acquisition:
Sensor (e.g. ultrasonic distance sensor)
RS485 (or similar) node

Hmm. Looks about the same, but you can sample the data as often as you like. Don’t forget, there are strict limits on the amount of data you can transfer with LoRaWAN. Everything is wired, so easy to troubleshoot. RS485 in noisy factory environments is very common and known to work.

Again, I suggest that LoRaWAN is not what you want. It could be that RS485 multidrop networks is not what you want either, but in my opinion a wired, powered, solution would be more appropriate.

Or use WiFi …

Like @ame I would be tempted to go industrial network & ‘old skool!’ Like RS485 or poss ind e’net, but in this case I would definitely look to LoRaWAN…

Readily available ultrasonic level sensors means little dev activity needed, depending on site conditions and silo infrastructure and other factors means likely 2 poss 3 gw’s would give redundancy of coverage and again quite low cost even if deploying dust rated devices. But for me there are 2 killer reasons to go wireless. Depending on where in the world the cost of pulling protected (in conduit, cable trays, and gantries, plus any needed local extensions) cables, including fact copper isn’t cheap and is even rising now economies emerging, will likely exceed cost of full LoRaWAN system. Assume sensors for either system likely similar cost. Gw vs plc cost? Similar?. 2nd item is once LoRaWAN network in place what else can be done with it? It’s the future proof option… client decides they want to monitor local (1 or 2 positions?) conditions in silos say Temp & Humidity of nuts… simples deploy selected T&H sensors as needed… very low cost option. Hummmm big chunky motors on each silo…fans or movement?..want to retrofit bearing temp sensor or vibration sensor for (predictive) maintenance? Easy add sensor, no new cabling needed. Place a few Bluetooth beacons around the site and have staff and large assets carry BT to LoRaWAN tx’ers and you have a site wide location system. Monitor general factory environment… t,h, lux, etc… select and deploy sensors as needed. And so on… all building off initial infrastructure… if concerned on capacity or coverage add additional gw’s and densify the network as and when needed all at low cost. That factory lighting @TonySmith mentioned… again add LoRaWAN lighting monitoring for fault detection luminaire efficiency monitoring, lamp maintenance. Concerned about kit around the being left on and wasting energy around the site ? Add simple power metering or monitoring (full meters or simple Ragowski coil sensors…) with operator or management alerting. Just let your imagination rip…

Have seen many similar deployment use cases leveraging LoRa/LoRaWAN, which far from perception isn’t ‘new’, LoRa as a technology was publicly announced 8 nearly 9 years ago now and LoRaWAN was publicly announced over 6 years ago. I saw my 1st test silo deployment (grain per @TonySmith ) over 7/years ago…and system has expanded substantially since then and still going strong. Btw initial update rate was deemed too slow after 6 months of testing…so in that case they went back and added a 2nd sensor to each site to effectively double update rate…and provide some system redundancy…so data rate limits called out…aren’t limits if you think laterally :wink:

One final thought wrt wireless a decade or two back this would have been a challenge with all that deployed metalwork, think reflections, multipaths and faraday dead spots with e.g. fsk or other legacy radio technologies… with spread-spectum LoRa has high degree of resilience and the reflections can even help cover dead spots in my experience.

This is the kind of site my ‘consultancy’ head loves to get its head around and deploy against…go for it…

Grab a couple of low cost gw’s plus say 10off Dragino LDDS75’s or equivalents from dozens of other vendors - e.g Decentlabs, a few t&h sensors (Laird RS186’s or Dragino LHT65’s or equivalents from dozens of vendors once again), and go test/play and see what works… you could be seeing data before 1st cable run even deployed… :slight_smile: :+1:

Just my two cents worth. (Ok biased as I’m an evangelist for the technology but in prior lives I have deployed wired systems and even, in intrinsically safe and in cases with long runs required, fibre optic based systems… and find wireless liberating and flexible…)

1 Like

I don’t disagree, and you make a compelling argument, however I think it’s worth pointing out that LoRaWAN may not be the solution, WiFi may not be the solution, and RS485 may not be the solution.

Having said that, since OP asked what hardware they might need to make something work I can tell you what we are using for water tank monitoring.

Basically, we have an ultrasonic level sensor at the top of the tank ‘looking down’. It has a 5 metre range and a 4-20 mA output. The sensor is powered from 12 Vdc.

The sensor is connected to an Ursalink UC1122 analog device. The analog inputs accept 4-20 mA signals and send data via LoRaWAN every 20 minutes. The device is powered from 12 Vdc and is located very close to the sensor itself.

In the case where there is no LoRaWAN gateway nearby we install a Mikrotik LoRa 9 kit (there is a LoRa 8 kit if you need other frequencies). The gateway needs power from somewhere, and and internet connection from somewhere. If there is no internet nearby we install a cellular modem. If there is no cellular service we get a station wagon and fill it with tapes.

The UC1122 sends its data. LoRaWAN magic happens, and the data appears in the console. I set up webhooks so that when the data arrives an HTTP call is made to a PHP script on a server. The server decodes the payload from the incoming data and stores the result (with a timestamp) in a database. Once the data is in the database you’re basically done.

There may be other ways of doing precisely this. In fact, I’d like to know, in case there is a better way than what we are doing.

I can see pros and cons of wired vs wireless for this particular project, but it is worth considering them so that your final decision is made with sensible choices, not just whatever is fashionable at the moment.

1 Like

200 silos?

Wired on a couple of close to hand ones, Wifi on a couple within range of a router, LoRaWAN on a couple well away from the infrastructure. Evaluate. Add some more of what’s working well. Evaluate.

Do not settle for a solution until you’ve evaluated - especially the capital costs & ongoing costs for which you’ll need Excel in dangerous mode to factor in cost of money (ie interest rates).

1 Like


dust can also produce explosive environments. At least when working with flour, you have to regard extra regulations regading explosion prevention.
In Europe, the common standard is ATEX. Research if you need ATEX certified hardware or not in your silos. This would make things way more complicated no matter what technology you use.

Best regards

Could not agree more. That’s why I suggested the first step is to ask the customer the type of protection required. DIP, Exe and Exi etc are completely different requirements to IPxy.

This is about limiting the energy when there is a technical problem with the electrical/electronics and secondly in the case of DIP how the temperature of the exhaust gas from a flame is reduced below the ignition temperature of the dust.

I can only suggest Googling silo explosions and how a small ignition knocks the combustible dust from the frame of a structure which then leads to an ever increasing explosion. The result is spectacular.

Thanks for the help ame i apreciate your feedback


We have just recently launched a new Stock Level Monitoring Sensor. It might be a good solution to your problem. Please have a look at our website:
Self-powered, LoRaWAN level device

You could also consider more industrial solution based on several RS485 sensors connected to a central unit that can stream data over Ethernet or WiFi.
Stock Level Monitoring System

So many ideas and I totally agree it’s just a matter of deploying a test scenario witch client is interested in doing.
I will keep researching and keep the post alive with updates.
Appreciate your message Jeff!

Cool , send me a DM

Too all your suggestions about fire related issues nuts are previously watched and the hole plant is open air , I can understand with grain this is a big issue, I don’t think would be here

Very cool , where do you get 12v from ? PSU , battery or maybe solar?

Those are Fans with gas heaters to dry the nuts

Which country news feed should I subscribe to, awaiting the sad news of the huge fire in the nut factory?

Silo’s are enclosed spaces, dust can accumulate, standing in front of the camera crew and saying “I didn’t think this would be an issue” will be a low point.

So I’d strongly recommend you find out the facts rather than assume.