Zoom Logo

WILDLABS Virtual Meetup: Drones - Shared screen with speaker view
Stephanie O'Donnell
26:03
Hi everyone! Welcome to our drones virtual meetup. As always, use the chat to say hello and introduce yourself, where you’re from and a bit about your work or why you’re interested in the topic.
Lenny
29:00
Hi everyone, my name is Lenny, I am calling from Tasmania and for my PhD, I am using drone based LiDAR to look at the relationship between forest structure and biodiversity.
Robin Sandfort
29:03
Hi I´m Robin (www.capreolus.at) working with wildlife related projects mainly in Austria and Germany. I use UAVs to monitor beaver and wild boar damages (automatic detection and measurment) and try counting roe deer.
Thomas Lewis
29:19
Hi all. I'm a PhD student from the University of Sheffield (with the Macaw Recovery Network & Chester Zoo) studying the ecology of the Great-Green macaw in Costa Rica. I am interested in using drones to survey for nest-sites in emergent trees, to help with a study into nest-site selection.
Kasim
30:17
Hi all, I'm Kas. Just finished up a PhD on leopard ecology and conservation, and generally just interested in wildlife tech.
Éadin O'Mahony
30:41
Hi! My name is Éadin (pronounced Aid-een, it’s Irish). I’m from Ireland, and work for the North Coast Cetacean Society (bcwhales.org) in Gitga’at native territory, British Columbia, Canada. I’ve just finished up 6 months of field work out there, flying drones using LiDAR and collecting blow samples from fin whales and humpbacks for my Masters. I’ll be developing the drone work on behalf of the organization and am excited to learn more from this webinar!
Calin Ardelean
30:53
Hi everyone, I'm Calin from WWF Romania, and I recently started to use a drone for wildlife monitoring so I am very interested by the subject.
Richard Milton
31:47
Hi, I'm Richard from CASA at University College London. We used drones to map Lima and also run education projects with kids
Talia Speaker
31:51
Hi all, thanks for joining! If you’re having any technical difficulties please feel free to message me privately as well.
Alejandro Tello Martinez
32:27
Hi everyone, I'm Alejandro Tello, a hydrogeologist and photographer with a vast experience in drones, from Spain now based in Lima.
Sydney Kaplan
33:33
Hello, I am Sydney and I am a masters student at Illinois Institute of Technology in Chicago. My research involves using drones to monitor tree health and take samples from the tree canopy.
Langen Mathew
33:43
Hi all, Im Langeni Mathew, GIS Specialist, WWF-Tanzania, Support all programmes
Mariella Leo
33:48
Hi all, I´m Mariella Leo, a biologist from Perú. Now in the city of Chachapoyas
Chris Marsh
35:48
Hi all, I'm Chris, a Ecologist from the UK, working in New Mexico and Sumatra on vegetation regrowth
Daniel Fitzgerald
36:23
Hi, All- Dan from Cellular Tracking Technologies. We are experimenting with our radio telemetry system, LifeTag, on drones.
Mazidi Abd Ghani
36:31
Hi I’m Mazidi from WWF-Malaysia, based in Borneo. I operate conservation drone for the organisation for various conservation activities including orang-utan conservation, wildlife monitoring and enforcement, aerial mapping and habitat restoration.
Stephanie O'Donnell
40:22
David Lang (OpenROV) spoke at our open source meetup last season.
Stephanie O'Donnell
40:50
https://www.wildlabs.net/resources/community-announcements/wildlabs-virtual-meetup-recording-low-cost-open-source-solutions
Lot Amorós
42:43
HI, I’m Lot Amorós, founder of the project Dronecoria I developed an open source drone for aerial restoration with aerial sowing after wildfires.
Lot Amorós
42:46
http://dronecoria.org
Thomas Starnes
47:24
Hi, I'm Thom Starnes, Senior GIS Analyst with RSPB (Birdlife in the UK). We areusing drone imagery to create aerial surveys for our 200 reserves and produced the Drones for GIS Best Practice document on WILDLABS Drones group https://www.wildlabs.net/community/thread/473. Great to be here.
Zoe Jewell
49:40
Hi I’m Zoe Jewell from WildTrack, a non-profit focusing on non-invasive wildlife monitoring. We have been using drones for local reconnaissance in Africa to find animal trails, and are interested in developing AI to have a drone identify trails and follow them.
Brian Chilambe
52:53
There are regulations in Zambia. Our drones project was used by the Zambia Civil Aviation Authority to implement and refine the governments regulation around drones which as well includes regulations to fly beyond visual line of sight. search for the "ZCAR part 18"
Stephanie O'Donnell
54:00
I think it might be worth starting a separate discussion about regulations on WILDLABS - the questions we had coming back we rather specific (e.g. ‘FAA requirement, Part 107?’)
Robin Sandfort
54:15
flying beyond visual line is not possible in Austria.
Stephanie O'Donnell
54:15
This is my favourite slide to date.
Alejandro Tello Martinez
55:48
Robin, same case in Spain. you are not allowed to fly a drone beyond the visual line
Tom Quigley
01:04:46
Is the machine-learning thermal IDing live on the drone flight itself, or post-processing?
Robin Sandfort
01:05:28
Thanks for the talk!
Biraj Shrestha
01:09:32
Nice work Claire! Did you also work in Nepal? I think I saw it on the map. This would be a great tool to do mega fauna census.
Biraj Shrestha
01:09:48
sure
Mazidi Abd Ghani
01:09:56
I have a question
Biraj Shrestha
01:10:00
yea please go ahead
Yvie ARMSTRONG
01:10:05
awesome work
Mariella Leo
01:10:36
Hi, may you comment on yor AAF project in Perú?
Biraj Shrestha
01:10:37
cool
Robin Sandfort
01:10:40
Do you have experience in finding deer fawns in Meadows?
Ned Horning
01:12:17
Can you explain what you mean by thermal profile? These look like single channel images. How do you extract thermal profiles?
Talia Speaker
01:12:35
Here’s that Orangutan Nest Watch: https://www.wildlabs.net/resources/case-studies/drones-orangutans-and-strangler-figs-malaysian-borneo
Talia Speaker
01:12:44
And Sol’s profile: https://www.wildlabs.net/users/sol-milne
Thomas Starnes
01:13:34
Fantastic presentation Claire! I wonder are you doing any development with multispectral / hyperspectral?
Joanna Turner
01:22:59
@Ned convolutional neural networks can learn patterns in single channel images. I think that might be how they do it.
Claire Burke
01:24:20
Replying to Ned andJoanna…The thermal profile is essentially the pattern of warmer and cooler spots on the body of animals. Whilst the absolute temperatures of these vary from animal to animal the pattern is quite consistent for each species. The images I showed are radiometric - so temperatures are real, but for ML we convert these to single channel RGB where temperature maps to the colour values
Ned Horning
01:26:09
Thanks - that makes sense. When you contrasted shape and thermal profile I was thinking you were recording multiple wavelengths
Joanna Turner
01:26:39
Thanks Claire
Claire Burke
01:27:12
The hyper/mutli spec work is for detecting chemicals in the environment, e.g. from disused mining sites. We couldn’t afford to but one ourselves, but we’ve been using the same tech in astro for centuries to determine the composition of stars. So I wondered if we could build our own, starting with the physics and focus on detecting the specific things we were interested in. I have a Phd student working on it now and so far things are progressing well
Stephanie O'Donnell
01:31:52
@eadin - just had a thought. If you haven’t already, you might chat to the Duke Marine UAS team, they’re doing so much work with marine mammal monitoring + drones. They might have some experience to share if you’re not in touch already?
Stephanie O'Donnell
01:32:07
https://twitter.com/MarineUAS
Claire Burke
01:35:15
Do you work with the local authorities in Madagascar on developing drone regulation and infrastructure etc?
Talia Speaker
01:37:12
Here’s the group Steph just mentioned: https://werobotics.org/flying-labs/
Éadin O'Mahony
01:38:58
@Stephanie I’ve not been in touch with those guys, I’ll definitely follow that up! Thank you!
Robin Sandfort
01:41:01
Are there ideas to join data like in the camera trapping community?
Robin Sandfort
01:41:32
Just like: https://www.wildlifeinsights.org/
Mikheil Potskhishvili
01:42:26
.
Stephanie O'Donnell
01:43:05
Great point robin.
Stephanie O'Donnell
01:43:09
Will throw it in there.
Éadin O'Mahony
01:46:34
The link to that paper would be fantastic please!
Stephanie O'Donnell
01:46:48
We will track it down!
Éadin O'Mahony
01:47:01
Thank you!
Stephanie O'Donnell
01:47:02
Our notes + recording will capture all of this info
Brian Chilambe
01:47:03
Thank Claire can you share you paper on how to get started with thermal cameras on drones
Thomas Starnes
01:48:25
Have you seen OpenAerialMap? https://openaerialmap.org/
Thomas Starnes
01:48:51
Upload all of your RGB (and maybe thermal?) imagery, view on global map and digitise into OpenStreetMap
Robin Sandfort
01:49:26
Thanks for the info Thomas!
Tom Quigley
01:50:20
+1 to this questions!
Talia Speaker
01:50:58
Here’s that paper from Claire! We’ll also link it in the notes we’ll publish with the recording https://www.tandfonline.com/doi/abs/10.1080/01431161.2018.1558372
Tom Quigley
01:51:29
What’s the most exciting thing (hardware, software, or ML) on the horizon that would change the game for drones? (Other than changes to flight rules like line-of-sight)
Éadin O'Mahony
01:51:33
Thanks Talia!
Lenny
01:52:28
Check Lian Pin Koh from Conservation Drones talking about drones and hype cycle
Lenny
01:52:29
https://www.youtube.com/watch?v=8ta-FwwwXxE
Éadin O'Mahony
01:53:09
A huge limitation for us has been the noise from the drones! Although definitely more non-invasive than traditional survey methods, we still get responses from animals to the presence of the drone
Stephanie O'Donnell
01:54:01
Oooh Eadin, that is a big area we need to delve into. There are conversations on wildlabs about this
Claire Burke
01:54:48
@eadin totally agree - we’re always doing our best to plan drone observations for minimal disturbance, but learning what works is a field in itself
Robin Sandfort
01:54:49
Good Point Eadin! It might also be the form of the flying object. Chamois reacting to golden eagle shapes.
Éadin O'Mahony
01:54:49
Smaller drones are quieter, but then you have shorter flight times - so big trade offs either way
Claire Burke
01:55:08
using fixed wing over multirotor is a good solution to this if you have them
Gwénaël DUCLOS
01:55:35
Do you think it will be usefull in the future to have several cameras on the same drone (for example: thermal + LIDAR or other couple) ?
Robin Sandfort
01:55:50
Fixed wing might be even more problematic for some species.
Éadin O'Mahony
01:56:10
Do you not need a large take off and landing area for fixed wings though Claire? I launch ours drones from a tiny research vessel and catch them out of the air directly.. (we have been using DJI magic 2 pros)
Russ Van Horn
01:56:30
One adaptable mammal (American black bear) may, or may not, habituate to overflights of drones: https://doi.org/10.1093/conphys/coy067
Robin Sandfort
01:57:05
One of the rare papers on that Topic.
Andy Quitmeyer
01:58:14
are tiny blimps viable at all?
Stephanie O'Donnell
01:58:33
https://www.wildlabs.net/community/thread/432 - this conversation might be one we should revise
Talia Speaker
01:58:34
Here’s one of those WILDLABS threads on drone impacts on wildlife: https://www.wildlabs.net/community/thread/432
Stephanie O'Donnell
01:58:43
Ha ! Thanks talia
Claire Burke
01:59:14
@eadin - you can hand launch some fixed wings, you sort of throw them at the sky and the fly. If yo have a waterproof body you can water land it then just scoop it up from the boat
Andy Quitmeyer
01:59:22
yay open source :) ❤️
Thomas Starnes
01:59:30
Agree with above comments about wildlife disturbance - we need to be really cautious. Encourage drone users to share their experiences, ideally through peer-reviewed literature but even anecdotally e.g. on the disturbance thread on WILDLABS Drones group will help to prevent us making the same mistakes
Thomas Starnes
01:59:44
Approaching birds with drones: first experiments and ethical guidelineshttps://royalsocietypublishing.org/doi/full/10.1098/rsbl.2014.0754
Éadin O'Mahony
02:00:10
@Claire wow ill have to look into this!
Tom Quigley
02:00:13
There are some military-grade quadcopters in development that are interesting as well - but cost prohibitive probably for anything but the biggest budgets
Mazidi Abd Ghani
02:00:31
In our case, smaller drone may cause some risk of bird attack. One of my drone was chased and attacked by eagle. Bigger drone will prevent this. In addition to that, I set my flying height at 200m where I found it safer based on bird of prey flying behavior
Tom Quigley
02:00:32
https://hrobotics.eu/
Andy Quitmeyer
02:02:27
I found it surprising in Panama how much the animals seemed not as much affected by the little dji mavic flying around, maybe they are used to the loud cicada sounds
Tom Quigley
02:02:48
Also a model I love in development is https://www.flyzipline.com/ - a solution looking for a problem that would fit it in the conservation world
Talia Speaker
02:03:27
Thank you all for joining! You can continue the discussion on WILDLABS here: https://www.wildlabs.net/community/thread/809
Chris Marsh
02:03:30
If anyone is interested in drone ecology work - https://chrisdmarsh.wixsite.com/dronecology
Alejandro Tello Martinez
02:03:33
thanks to all of you!
Andy Quitmeyer
02:03:35
thanks for getting us together!
Talia Speaker
02:03:50
Next meetup will be on eDNA: https://www.wildlabs.net/resources/community-announcements/wildlabs-virtual-meetup-invitation-environmental-dna-edna
Éadin O'Mahony
02:03:56
Thanks so much! This was wonderful!
Tom Quigley
02:03:56
Thanks all & thanks Steph for hosting!
Thomas Starnes
02:04:07
Thanks everyone!
Mazidi Abd Ghani
02:04:09
bye
Calin Ardelean
02:04:13
Thanks a lot
Robin Sandfort
02:04:20
Thank you everybody!
Mikheil Potskhishvili
02:04:23
thank you