Good day
reading through the forum on regulations in SA, is my understanding correct.
We have a 1% duty limit which is around 14.5 min a day.
If my device needs approx. 200ms to send the data, I can only send ~75 messages a day which makes it not a fit for a tracking device which needs at least an update every 5-10 min which would be 144 messages a day
Thanks!
Search the forum and youâll find numerous discussions about why LoRaWAN is not suitable for active tracking - itâs not a design flaw, itâs just not an appropriate use case, mostly as there isnât the sort of coverage that you get with mobile/GSM/LTE etc
âIâve arrivedâ or âIâm hereâ tracking is feasible with care.
Thank you for the fast answer, really appreciated it!
I have an use case in South Africa where there is no mobile network at all available for kilometers but I understand LoRaWAN is then not the alternative
Thanks again
Its quite likely in a tracking application that the tracker can find itself a long way from a gateway. So the tracker goes to long range mode and each location message takes around 1400mS to send, so your limited with TTN to 21 messages a day, an interval of about 70 minutes.
Strictly speaking, itâs any radio in the 868-915MHz ISM band thatâs not got the duty cycle you are looking for, the protocol isnât relevant.
What is it that you need to track that requires an update every 5-10 minutes thatâs so far from civilisation that there is no mobile coverage? Maybe we can point you to a solution - like Iridium if the tracking value is sufficiently important to warrant the message cost.
You may need to get some new batteries for your calculator as well, as re-looking at your first message youâve divided minutes by milliseconds - the actual answer is 4320 messages at 16 bytes each assuming you have a set of gateways or, not for discussion here, point to point LoRa.
If you can give us more info, we may be able to help!
1 Like
you beat my calculator, thatâs for sure. We have areas in South Africa (and other African countries) where you donât get a mobile connection for kilometers. The aim is to track an animal. Elephants and Rhinos to be precise. Not fast moving so 5-15min is ok. The park size is ~5x5km but a coverage of 10x10km would be great in case of run away or theft.
Rather than re-inventing the wheel, or breaking your calculator, you might want to review some of the existing solutions exactly for that implemented by the likes of ZSL, IRNAS, and others for anti-poaching measures and general endangered species protection - There have been several overviews presented at e.g. TTConfs in the pastâŚ
1 Like
Whatâs wrong with the products already on the market? (Genuine question, if there is a flaw in them, mainland UK only has a few elephants & rhinos so Iâm not up to date on such things).
As for tracking, do you really need to know where they are every 5-15 minutes? I mean really really need to know?
How about updates when they move &/or stop? And lots of updates if they suddenly start moving fast (for them, as in being chased) and updates if they are suddenly moving fastish in straight lines (ie on a vehicle). As well as being geofenced - so hardly any updates whilst in their normal spots / centre of the park and more when closer to boundaries or outside of normal haunts?
There are also herd aspects - devices can talk to each other so if any one signal gets through, it says whoâs with them - so they can all stagger their transmissions to spread the battery load - not yet seen a solar panel mount for a larger animal but I guess itâs feasible to do a top up of the battery - otherwise 5 - 15 minute updates will require them to lie on a charger pad at night or youâll have to change the batteries every couple of weeks or teach them how to plug it in.
I did write an article describing a simple way of searching for GPS trackers over an area of 500sqkm in 10 minutes, but its point to point LoRa and thus off topic on this TTN forum, so I should say no more.
Get up high!
I see some UAV experiments in the future - not here in the UK, that would be wrong / illegal / naughty âŚ
Well exactly.
With legal restrictions on power output levels, if you want more range to receive stuff, then you need to turn the reception path into line of sight with height.
Have a look at this calculator.
https://avbentem.github.io/airtime-calculator/
You can get about 500 tracking messages through per day onthe 1% duty cycle - a tracking message is typically 10-20 bytes long. Can be shorter though you wont get mucjh better,
My trackers work easily for 5 - 50 Km depending on line of sight
LoRa is great for this type of tracking.
Iâd be glad to chat to you - I live in Joburg.
1 Like
But donât try that kind of volume on TTN as you will exceed the fair access policy.
Btw, you assume SF7 with those numbers. Is that realistic in an area where (I assume) there will be trees hills and the like?
Seems there is a general calculator problem in the Southern Hemisphere - on SF7 15 bytes you can do ~500 messages per HOUR on 1%. But as @kersing notes, SF7 is hugely optimistic for even 5km, let alone 50km.
thanks for the link! Even 36 messages per hour with 20 bytes payload and SF11 would be more than enough for my use case. Thanks again
Remember that is under the 1% duty limit - not on TTN which has far tighter Fair Use limits. Also remember, RF spectrum is a scarce shared resourceâŚjust because legally you âcanâ does not mean in practice you âshouldâ! If every user took the view âhey I can max out the 1%â we would quickly get congenstion and network data failuresâŚplease think how your application might work at maybe 10% or even 1% or what is legally allowed⌠appreciate there may be a âpresumptionâ of is in the wilds/in the savannah or whatever and surely not many other users around but you never know and that very environment is likely to be a driver for other LoRaWAN users. Also its in the name - Long Range WAN - meaning your excess transmission may go a long way and impact other distant users 
Also check what the actual payload of your tracker is and what data is sent - many I have seen donât just send location but will likely send other âusefulâ data even battery level, temp, barometric pressure, motion/orientation data etc. adding to payload size. Even decimating location resolution using e.g. Cayenne LPP formating have seen trackers with 21 to 27byte data sets, which can further limit message rate c/w your expected15-20.
Probably battery lifetime is an issue too? I think common GPS trackers work for days or weeks, but probably not months or even years.
Maybe lorawan geolocation could work for you? I am not completely up to date, but I think Kerlink is producing gateways with inbuilt geolocation functionality. This way you could use a low-power (and low airtime) lorawan node placed on your animals. Donât know the minimal number of required gateways for good geolocation results though.
But this would be no longer TTN related.
Why not? TTN has a LoRa Cloud integration that allows turning on geolocation services.
Youâd be thinking of Time of Flight which requires an upgraded concentrator module = ÂŁ/$/âŹ1200+ and relies on three or more gateways to hear the uplink. The poachers will probably just steal the gateways.
But there is the all new LR1110 module that hears the GPS and sends the data to a Semtech API that interprets it to provide low cost low power geolocation.
But youâd have to check that GPS covers that area âŚ
Oh, thanks for pointing this out, @descartes and @kersing
Seems the LR1110 could be a very interesting module!
Yes, geolocation via ToF is expensive⌠the ToF capable Kerlink Gateway is around 1â500 EUR in my region and you need at least 3 of them.
With default gateways you could use the WisBlock Kit 2 GPS Tracker which seems to be equipped with a solar panel.