Member Spotlight: Wildlife.ai

We are thrilled to have Wildlife.ai in our second cohort! As we all know and understand now, the climate and biodiversity crises can't be tackled as separate things.

Wildlife.ai uses artificial intelligence to speed up wildlife conservation. It fast-tracks data processing by integrating citizen science and machine learning in conservation monitoring. 

A better understanding of ecosystem dynamics enabled by Wildlife.ai's tools accelerates restoration and protection to mitigate the devastating impacts happening to our natural environment. These tools provide solutions to better manage and conserve our wildlife and biodiversity. 

During this segment, we spoke with Victor Anton, the founder of Wildlife.ai! We discuss the many projects at Wildlife.ai and how one can get involved and delve into AI and its many uses. 

1.Tell us a bit about your background and what inspired you to start-up Wildlife.ai.

I am a wildlife biologist by training and am originally from Spain, where I completed my Environmental Science degree. I was really interested in the subjects related to ecology and conservation. I decided that learning about one ecosystem was narrowing down my skill set so I wanted to see how ecosystems work around the world. I was lucky enough to go to Norway and learn about alpine ecology for a year. Then I moved to New Zealand to do a Masters in Ecological Restoration and learn about temperate rainforests in Wellington at Victoria University. 

That enabled me to see how important the community’s role is in conservation, how important it is to monitor the environment so you can see what is happening, and how you can manage it to protect the targeted species. After my Masters, I worked in a research position looking into invasive species in New Zealand, such as rats and possums, which led me to my Ph.D. thesis at Victoria University. In my thesis, I researched how populations of invasive species adapt to urban environments.

For that research, I used motion trigger cameras, collected thousands of images of animals, and needed to analyse them. I thought it would be good to share the task of finding out what kind of animals are around in people’s backyards and so, I engaged with the community to speed up the task of processing the photos. In this process, I became more familiar with citizen science, how you engage people in scientific processes, and how it can be integrated with machine learning and AI.

2.What are the current projects at Wildlife.ai and how can people get involved?

People can get involved just from their couches! They can save species from the comforts of home by becoming citizen scientists and helping classify some of the data that researchers or conservation biologists have collected.

Anyone can also become a ranger at Wildlife.ai! We have many different projects and rely on rangers to collect and analyse the data; Wildlife.ai rangers are volunteers passionate about the environment that want to contribute, with their skills, to accelerate species conservation.

One of our projects is Spyfish Aotearoa, which is in collaboration with the NZ Department of Conservation. In Spyfish Aotearoa, we better understand the biodiversity of Marine Protected Areas and how it influences the coastal ecosystem. We do this using Baited Underwater Videos and recording fish counts inside and outside of the reserve over time. There are thousands and thousands of videos collected, and we make them available for citizen scientists to discover what types of animals live in the waters in NZ and how they can be involved in that monitoring from home. 

Another project is 'Wild about AI', targeted at 9-11-year-old students where they can learn how to create their own artificial intelligence models and identify fish species in their local environment. Wild About AI uses different online resources, including Spyfish Aotearoa, to expose students to science in a fun and engaging way. This resource is great for educators!

We’re also developing a camera to identify wildlife called ‘Wildlife Watcher’. This open-source smart camera trap gets triggered based on the movement of target animals, identifies the species, and reports the observation in near real-time to biologists, enabling more efficient species conservation efforts.

In addition, we’re working with the Israel Nature Park Authority (INPA) on a project to learn more about the declining population of Griffon Vultures and how to protect them. The biologists from INPA have a lot of GPS and accelerometer data of the vultures that we are integrating with machine learning to discover how we can better manage and prevent this species' extinction. 

3.How are you finding the uptake of machine learning and AI tools for research purposes?

Over the past three years, there has been a huge development and uptake of AI and, more importantly, an understanding of what AI is in the conservation community. There is a growing community, there's no doubt about that. Mainly because it is easier to gather and access large amounts of data, so machine learning is the perfect tool to analyse that information.

The transition from research-based machine learning projects to easy-to-use applications for the conservation community is still an area that needs to be improved. The thing that is still missing and what Wildlife.ai is working towards, is making that transition from research-based machine learning to more community application. We don’t want AI tools to only be used for research purposes.

We know research is important, and it is often the first point of access, but we would like to bring these digital tools to the community and people on the ground. That’s a big part of the work that we're doing!

4.What are some of the lessons you’ve learned from the research conducted at Wildlife.ai? 

A major one is that the most challenging part isn’t about the algorithms or how to implement AI. It is how to get the community engaged and clearly identify the local conservation challenge they are trying to tackle.

We have access to a lot of data, but it isn’t of much value unless we know what the real needs are. Digital solutions are often inefficient without fully understanding the local conservation challenge, the resources available, and how the proposed solution will be maintained and updated to future needs.

Common challenges associated with AI are also hindering a faster uptake of this technology in conservation. Currently, there is no clear path or guidelines for people to use AI tools. Everyone knows it is important and that they should be using it, but it is still unclear how to implement it for data processing and collation. As a result, conservationists start trying AI on their projects but don’t have the right resources and end up doing business as usual.

At Wildlife.ai, we ensure conservationists and data scientists codesign the projects, plan realistic roadmaps, and secure the right resources to successfully use AI for conservation. 

5.How has your involvement with Subak been so far and what are you hoping to get out of joining our ecosystem in the long run?

The involvement with Subak has been great! We’re learning a lot of different aspects about being a start-up and what you have to be aware of.

It’s not only about the finances or workshops on storytelling, legislation, etc, but it’s also great being part of a wider ecosystem. The connections are great by joining a community with the same values. It feels more natural to ask for support and advice when your values align. 

Everyone at Subak has the same values, making the road much easier to connect, support others, ask for favors or build partnerships.