article / 10 February 2022

Introducing the WILDLABS On the Edge Fellows

We're proud to introduce the first WILDLABS On the Edge Fellows for 2022, Loretta Schindlerova and Meredith Palmer! Working alongside expert Edge Impulse mentors, these two fellows will use embedded machine learning to bring innovative projects to life.

In 2021, WILDLABS partnered with the machine learning experts at Edge Impulse to launch WILDLABS Fellowships: On the Edge, the first in our growing fellowship programme

Fellowship main banner

With a $6,500 award, expert AI mentorship from Edge Impulse, and access to the world’s biggest conservation technology community, our first Fellowship set out to show how innovative embedded machine learning, one of the most promising and rapidly evolving conservation technologies, can be impactfully applied to real projects within the WILDLABS community.

Funding may be the fuel of conservation tech projects, but accessible technologies are integral to steering projects toward success. Providing access to tech mentors with expertise to share allowed us to open up applications to an exciting and extremely important section of our community: those with bold and innovative ideas, but without the technical machine learning skills to reach the finish line on their own. Alongside their mentors from Edge Impulse and the WILDLABS team, the chosen Fellows will work at the intersection of engineering and conservation to bring their innovative, cutting-edge ideas to life.

By combining the creative energy and field expertise of conservationists with Edge Impulse mentors’ deep knowledge of embedded machine learning’s potential, our chosen Fellows will have the opportunity not only to drive their current projects forward, but to make a difference throughout their careers with their new ML skills!

Meet the On the Edge Fellows

meredith

Meredith Palmer

Project: It’s a trap! Embedded AI activates modular components of open-source camera trap to capture novel ecological data

We will optimize an embedded wildlife-classifying AI on open-source camera trap technology to create a 'smart' wildlife  monitoring device that takes images, sends alerts, and gathers biological data when triggered by target.

Our goal is to combine our past hardware and software developments to create CTs that make  smart decisions about how they sample and react to the environment. Not only can we  overcome critical limitations by embedding deep learning at the CT level, but this application  also opens up new frontiers in terms of how we manipulate and study ecological systems. By year’s end, we aim to develop a microcontroller embedded with TinyML that interfaces with  our open-source CT modules and tech from other conservation technology companies.

A critical component of this fellowship will be to support capacity building efforts in WILDLABS’ East African regional community, including providing leadership and oversight to technical skill-building workshops for local conservationists, allowing the impact of this Fellowship to reach even further.

We asked Meredith…

Why do you think it’s important to share your conservation tech project and the process of being mentored with our WILDLABS community? 

As a conservation practitioner, I often find myself limited by my own lack of understanding in terms of the exciting possibilities offered by emerging technology. As we develop these new tools, sharing the process of discovery (and failure!) helps pave the way for others interested in tinkering with similar tech.  

What are you looking forward to about this fellowship, and what do you hope to get out of the next year? 

This fellowship aims to deliver impact on multiple levels: not only do we strive to produce new open-source conservation tools, but we plan to do so by involving and empowering local technological communities. I can’t wait to start expanding my own knowledge of TinyML and sharing the wealth of information and expertise provided by Edge Impulse with the East African AI community. Working collaboratively across these partners enables us to put the ownership and stewardship of conservation tech into the hands of those striving to save the ecological systems that impact them the most.   

Why do you think embedded machine learning (and machine learning in general) are so important to the future of conservation tech? And how can making these tools more accessible to conservationists change this field? 

We need to embrace out-of-the-box thinking for how we can adapt current technologies to collect novel forms of ecological data. Embedded AI represents an exciting new frontier for our conservation programs, enabling us to reframe current tools – such as camera traps – as smart, comprehensive biomonitoring devices that can interact with the environment in novel ways. Sharing our hardware and software with the conservation tech community promotes uptake of these developments while stimulating discussion on how we can improve or modify these tools to take on a broad range of conservation challenges.  

loretta

Loretta Schindlerova

Project: Real time acoustic localization of wolves for research  and conservation

We will be deploying a prototype network of acoustic sensors that can pinpoint the location of howling wolves in real time, and thereby track the movement of packs to mitigate conflict with humans. 

Wolf-human conflict is a major threat to wolf conservation and ecosystem restoration. Detailed information on pack movement and individual dispersal is currently limited because of the high cost of individually trapping and GPS collaring wolves. Additionally, in even the most intensively studied area, most packs contain no more than one or two individuals with a collar, providing very little information on intra-pack dynamics. Wolf howling plays a key role in pack cohesion and movement, and therefore knowing where and how all the pack members are howling and responding to howls can give indication of movement, habitat use, hunting  behaviour, and dispersal of young animals from natal packs.

I would like to see the working system freely and openly available to researchers not only working with canids, but with other species with long-range vocal  communication as well.

We asked Loretta…

Why do you think it’s important to share your conservation tech project and the process of being mentored with our WILDLABS community?

My project involves a unique, precise, and non-invasive method of animal localization. If the processes required for it can be improved using ML, it would undoubtedly become a valuable tool to many other researchers as well. I am sure that sharing the progress on this platform could be very beneficial for the community as a whole.

What are you looking forward to about this fellowship, and what do you hope to get out of the next year?

I am looking forward to connecting with and learning from like-minded people who might be using similar technology. I am also very excited about the field work and hoping to get to a point where it would be possible to test the developed system in the field.

Why do you think embedded machine learning (and machine learning in general) are so important to the future of conservation tech? And how can making these tools more accessible to conservationists change this field?

In my experience with the study of animal behaviour, and particularly bioacoustics, a very large amount of time is spent on purely manual work of just finding, selecting, and manipulating data. Using ML for these tasks would substantially improve the tempo of advancement in the field of animal ecology research. This in turn also influences our ability to react faster to rapid changes in the environment, which is crucial to the protection of various species and nature conservation in general.

The Mentors

mentors

Throughout 2022, Edge Impulse mentors Arjan Kamphuis and Daniel Situnayake will work with Loretta and Meredith to support their fellowship journey, provide guidance and advice, share their technical expertise, and help our fellows develop impactful projects and tools. 

And in turn, our Fellows will share their journeys with all of you in the WILDLABS community and beyond! In the coming months, you’ll get to know our Fellows and their work through interviews, events, project logs, and more, allowing other conservationists to learn from their experiences and engage in the Fellowship process! 

For more on Meredith, Loretta, and mentors Daniel and Arjan, stay tuned to WILDLABS for the latest news, and follow us on Twitter @WILDLABSNET for exclusive Fellowship updates!

Interested in learning more about our fellowship programme or building your own fellowship sponsorship package with our team? Find more information here.


Add the first post in this thread.

Want to share your own conservation tech experiences and expertise with our growing global community? Login or register to start posting!