Thanks Alasdair

The application I have in mind is for very precise tracking of terrestrial large carnivores and I would be doing ground to ground data downloads and settings uploads via UHF, with a GPS fix every 5 minutes that the animals are moving, reduced to one per hour when they are stationary - so the power drain is substantial. I am in contact with Open Collar, so we'll see what is possible within the constraints.

How are those thermopile imagers progressing - I do a lot of camera trapping and a sensor that was selective for real animals would be a huge advance.

Peter

Hello Peter!

We have been using small UHF based GSP collars with solar panels for arctic foxes in Greenland since 2017. Nothing DIY or open source - but developed by the polish company Milsar. 

Download range of these are a few kms.

Electronics and software works well but the ruggedness is still lacking. They have basically repurposed a bird tag but have little experience with making collars.

The huge advantage with the solar panel is that we could log at very high frequency - during summer time! 2-5min intervals even with these tiny collars weighing about 100g.

German company e-Obs makes rugged UHF GPS collars but they are not using solar panels. They have plans on using IMUs and magnetometers on future models - allowing for dead reckoning - alleviating the need for high frequency position logging a bit.

I have seen a solar GPS UHF collar from Italian company Technosmart but it is not on their web site:

https://www.cheetah-research.org/technical-developments 

https://twitter.com/technosmarteu/status/1192819618517307393

This last one seems like it could fit your bill - but it is not LoRa based.

Best regards,

Lars

Glad to see this type of tech is miniaturizing! I work with ruffed lemurs (~3kg), and the majority of the rest of lemur species are <1kg. But those tiny ones are all nocturnal meaning human tracking is very inefficient, so tiny, decently-affordable GPS collars would be AMAZING. I am completely on the conservation side of conservation tech, so the specs of all this are foreign to me, but excited to be able hopefully try some of this stuff out soon!

Thanks Lars, I will see what Milsar has to offer. I was in touch with e-obs but they didn't have quite the combination of weight, lefetime and collar material that I wanted. 

The project has now leapfrogged past the need for a high frequency of GPS fixes, and so wil will be piggy-backing our tracking onto collars that othe researchers are putting out.

Peter

You are welcome, Peter!

As I mentioned the ruggedness of the Milsar collars still leaves much to be desired. 
I would probably try contacting someone from the cheetah project linked above and ask the about their experiences with the Technosmart collars.

e-Obs provided the collars for the study of collective behavior of baboons so they should be rather rugged. Also, e-obs are going to integrate IMUs in the upcoming generation of collars. They are definitely worth watching carefully. They have no plan to integrate solar panels in their collar though.

https://e-obs.de/wp-content/uploads/2021/01/The-e-obs-Newsletter-2020.pdf

Best regards,

Lars

Hi Lars,

At the risk of corrupting this thread too far from the OP :

What components of the Milsar units are failing for you on the foxes and what fox behaviour is causing that issue?

We are currently deploying Milsar units on Koalas. They give their collars a fair work-out but they are holding together so far. Our biggest problem is keeping the GPS side of the collar facing up as koalas occupy a range of postures and sometimes the unit gets stuck between the koala and a tree trunk. This results in no charging and poorer satellite reception. We still manage 50 fixes a day with a modified firmware that improves accuracy but also increases power consumption.

Hi Matthew,

I am not entirely sure how the collars have failed since we loose them after failure despite a big effort in recapturing the animals. In one case where we did recapture, the UHF long range communication was gone but data was downloaded after the collar came in near visinity of the base station. Ecotone makes a similar collar with a drop-off mechanism but I have not tried it.

Thanks Alasdair

Could you encourage Heinmann to also test with dogs and cats - cattle and horses are right at the top end of the size range for camera trapping.

I am looking at LORA for the carnivore tracking - mainy to provice occasional real-time locations so that we can find the wild dogs to download stored GPS data and to do direct observations. The plan is to replace time consuming and expensive VHF ground tracking. I am concerned that nobody has used LORA on anything as small as an African wild dog and I am reluctant to risk our scarce resources on being the guinea pig.

Hi Peter,

The good news is that we (OpenCollar) are already miniaturizing the LoRaWAN + GPS technology to also work on the scale of African Wild dogs. Keep us informed on the specific needs and use-cases you know in the wild, so we can get the right configuration. We will update as soon as we have good info to share!!

Thanks Tim, our skype call gave me lots to think about. Juggling battery capacities and number of GOS fixes suggests that adding an extra C-size lithium cell may make more sense than using solar panels for charging, and would certainly simplify the technology and make it more robust.

You got me thinking, so I just ordered an LTC3588!  I figured, I have 2 dogs...