Current monitoring methods either don’t have the capacity to scale globally, or simply don’t have the required resolutions––and fine-scale data is often not within reach.
Copyright by Mark Minevich, www.forbes.com
The world’s biodiversity status is in crisis mode––and Covid-19 has only exacerbated this reality. Covid has served as a stark reminder that negative interactions with species can directly impact our lives. As of 1970, the world has seen a significant 68% average decline of birds, amphibians, mammals, fish, and reptiles. Just in the Americas alone, natural ecosystems provide humans an estimated $24 trillion worth of economic value every year, equivalent to the entire gross domestic product. As wildlife changes occur, all ecosystems become less resilient and are more at risk. Without resilient ecosystems, agriculture, water and wildlife-based tourism are left in significantly vulnerable shape.
Current monitoring methods either don’t have the capacity to scale globally, or simply don’t have the required resolutions––and fine-scale data is often not within reach. Governments and administrations have been slow to install measures; meanwhile, private sector employers desire the competitive advantages that come with ‘green’ credentials, but don’t always know how to contribute effectively, leading to green-washing and wasted resources. In addition, employees prefer to work for companies with a good environmental record and welcome a chance to volunteer and participate, but engagement is often symbolic or short-term.
Traditionally, researchers have been working tediously by completing manual tasks including identifying specific animals from photo shoots for population studies to classifying the camera photos gathered by field workers. We need to pool together a smarter global effort led by the UN, public and private sectors to bring about accurate, data driven current, global maps and hotspots of species numbers and distributions to develop prescriptive global conservation strategies. If we are to save our world’s biodiversity, now more than ever, it is time to mobilize fully-integrated AI and machine learning solutions for wildlife conservation––and to ensure that these solutions are sustainable for decades to come.
Unlike the domains of finance, science, healthcare and the like, wildlife conservation is often left in the dark when it comes to advanced AI solutions. Nevertheless, there are global pioneering organizations and startups that are working towards real use cases in AI for Good applications to bring about resilient biodiversity. For example, the World Wildlife Fund (WWF) is working with Intel to apply AI to monitoring and protecting Siberian tigers in northeastern China. According to the International Union for Conservation of Nature (IUCN), the South China Tiger is a “critically endangered” species. Intel’s Movidius visual device, combined with the company’s back-end analysis and recognition platform, are leveraged by the WWF to track the habits of tigers, collect data on them, and use this information to help restore their wildlife resilience. On the topic of visual recognition, as per Synced, “Although image recognition is the most widely applied AI tech in wildlife conservation, researchers and startups have also leveraged other tech to create devices and systems to protect animals in more proactive ways. PAWS (protection assistance for wildlife securities) is an AI tool designed to help rangers in the fight against poachers. It collects historical data of poaching activities and suggests patrol routes according to where poaching is most likely to occur.”
In addition to Intel corporation, companies like WildTrack are also pioneering data driven biodiversity solutions. According to WildTrack, the organization’s “AI-enabled Footprint Identification Technology (FIT) offers a cost-effective and non-invasive tool to collect, analyze and distribute data on species numbers and distribution at the scale and resolution required.” Moreover, WildTrack champions the approach of data democratization. According to the British Ecological Society, democratizing data collection to include environmental supporters is a huge unexploited opportunity. Ecotourists, local communities (especially those with expert indigenous trackers, e.g. current partners in Botswana, Germany, Israel, & Namibia), outdoor enthusiasts, schools and universities could collect data across borders. Because of WildTrack FIT’s interactive interface, the company has the ability to encourage direct engagement in conservation principles. This addresses the importance of interactive digital assets and tools in the creation of more robust ecological frameworks.
Lastly, it is important to outline that all applied AI solutions for conservation must stem from the core premise of sustainability. Sustainable solutions include the following: […]