Thx, ArjanS! I thought, Pet or Human tracking might be the application scenario. Given to the character of LoRa it s not suitable for high-speed tracking, hypothetically. Should I put my Qs there as well?
At the recent Lora workshop in Breda KPN made a presentation discussing the rollout of their dutch national Lora network. They said a big feature of their network is geo-location through triangulation based on time of flight using atomic clock accurate synchronisation between their gateways which are hosted on their mobile phone tower infrastructure. Their claim was 30 meter resolution which is pretty interesting but then your committing to their costing plan.
RSSI might be difficult due to the antennae orientation where actually the range of Lora works against you. However in sufficiently dense networks in shorter range Lora modes I imagine you could have some success with triangulation especially if the network server was involved cooperatively. Its something I am planning to research as it opens up interesting possibilities.
Great info! Very helpful! I am so grateful! Yes, you r right, I prefer TDoA to RSSI as well. But I didnt consider more details involving antennae issues like you that comprehensively, thanks again for the heads up. Frankly, the thing only that brothers me is the pre-geographic-measuring for RSSI. 'Cause I have no idea about it.
I am working on such tracker..
Thank you for interest.. first of all tracking via LoraWAN has number of limitations.. since the tracking should be near real time, the positioning data should be streamed every <15 sec.. this way one dog will use the same network resources as 100 sensors checked every 20 min..
Would it really need to be every 15 seconds? Every 2 minutes or even 10 would give you lots of positions. Otherwise as you say the data would start mounting up. I guess storing and forwarding all those 15 seconds points would also be too much. Perhaps if its for lost pets, when the pet is within a home geo fenced area the tracker could be quiet.
please, ask people who have lost their pets.. for two minutes a dog on run can pass two blocks.. any way, the scenario of pet-trucker should came upfront of technical specifications.. so called customer development..
And even when outside of a geo-fenced area, a smart protocol could drastically limit the data usage. Like see Best practices when sending GPS location data and Best practices to limit application payloads.
I have developed a product that was intended for agri-business (tracking cattle, research on animal migration, etc.)
They wanted the 3 axis accel for monitoring the activity level of the animal.
Still working out the final details.
so far.. (simplified feature list)
- LORA transceiver
- GPS (low power, multiple constellations)
- Cortex M0+ processor
- 4 Mb of flash (data logging)
- 3 axis gyro, 3 axis accel. 3 axis magnetometer
- Bluetooth (to allow direct access to smart phone/tablet/laptop..allowing a unit to act as a simple gateway)
- micro USB port (alternative power source for gateway, up/dn loading of code and data logging)
- CR2450 coin cell (650ma).. I expect better than 1 year battery life with reasonable reporting periods.
- Size: Polycase FB-45 (2.75" x 1.4" x 0.7".. keyfob case).. .this choice could easily be changed.
The design allows for nearly total shutdown and wakeup on timed basis or sensing of movement.
GPS idle currents are a killer (20uA).
Hardware design in place (not easy - keeping those antennas playing nicely together in small space!).. software still in progress.
Hard to judge the final performance (range of LORA) until everything is in place.
If "rechargeable" battery was required.. could be done.
But with 1-3 years of battery life, hard to justify it's value vs impact on costs.
It would make the obtaining an IP-67 rating easier..
Still haven't figured out best path forward for marketing for alternative uses (like tracking pets).
Open for ideas.. bit concerned about locking in a marketing path (hard to change later).
Brendan's observations are very astute. Great ideas? - easy. Getting heard - tough.
For most people during the next year.. cell coverage of LTE category 0 will be coming on. (much, much lower power requirements) will be impacting this market for several years. Would make battery life for the existing players MUCH longer. Harder to differentiate the value of LORA in urban settings.
Despite what the Cell providers tell you.. there are very large areas in the US not covered by them. Often, these places are where you can lose a pet while camping or hunting.
Mmm.. maybe I should be talking to "Bass Pro" or "Cabelas"?
Based on what I have seen of the existing products.. Dog owners concerned with tracking, often have multiple dogs.
Node ID is likely needed to allow tracking of multiple animals.
oops.. didn't read your original post correctly.. sorry.
(late in the day...eyes playing tricks on me)
As you're also logging the data: what are you using the LoRa for, and how much data is sent?
(It seems that a pet finder on The Things Network will require a good connection, as at low data rates one quickly hits the 30 seconds/day air time Fair Access Policy limit.)
We have come a long way with making our LoRa PetFinder, which we call CatTracer. See here our first very small GPS LoRa tracking prototype which posted its data nicely as we wanted it:
It seems - for urban settings- there is a LoRa based tracker you can purchase.
iotatracker.com ... $149
includes a "home base station" aka gateway.
small.. so it appears to have a limited range (chip antenna in unit that would be installed on collar).
another one that uses LoRa and GPS..
the home base (gateway).. doesn't appear to share gateways but has a bit better range on the radio.
@UncleJohn, just for my curiosity - is IOTERA using LoRa? They claimed that they have their own wireless technology which is patent pending.
IOTERA said during the kickstarter campaign a year ago it is using LoRa but later removed that comment.