Artificial intelligence meets the jungle: can AI help to protect our rainforests?

Massive deforestation worldwide is threatening the world’s ecosystems, climate, and biodiversity and, in the long term, our food supply. Global Forest Watch estimates that forests were reduced by 488 million hectares of tree cover worldwide between 2001 and 2023, and that pace is accelerating. A great deal of effort has been made by NGOs, governments, local communities, and many other stakeholders to reduce the pace of deforestation, particularly of rainforests, and investments have even been made to partially reforest the land. Unfortunately, the results of these initiatives are often not encouraging; across the globe, commercial groups and firms engaged in both legal and illegal activities are winning, while indigenous people and our natural habitats are poorly armed and ill prepared to defend themselves. It remains to be a very complex situation, where simple solutions are not on hand.

 However, cost-effective technologies, such as intelligent data collection through sensors or drones, can help gather critical information about the health of the forests and can detect illegal activities early, allowing indigenous communities to drive offenders off their land—or allowing government forces and conservation officials to arrive much earlier. There are three key areas in which technology and AI can help to protect rainforests.

• Technology for detecting illegal deforestation

A study published in the journal Proceedings of the National Academy of Sciences (PNAS) found, based on research in the Peruvian Amazon rainforest, that rainforest loss from illegal activities could be reduced by as much as 20% across all indigenous lands. That’s a lot. For the study, communities of indigenous people were equipped with smartphones and satellite data to identify where illegal activities were happening and to make it possible to stop them, either by themselves or with the help of the authorities. 

While that study was based on relatively low-tech devices, the PNAS researchers concluded that the approach is effective if there are communities on the ground ready to respond to threats and if monitors are well-trained and incentivized. However, a project in Indonesia, where a third of global rainforests are located, went a step further. The key technologies went beyond satellite data to detecting even minor disturbances in the rainforests. Acoustic monitoring, drones, remote sensing, and geo-tracking were combined to facilitate more effective responses by informing responders in real time about where disturbances were occurring. Similar to the example from Peru, the monitoring and remote sensing technologies were used to raise the alarm so that authorities, conservation managers, and other groups could be alerted to the need for a closer inspection—for example, if the sound of a chainsaw were detected.

A report by the National Geographic Magazine, describes another important avenue beyond early detection of deforestation by satellites, drones or smart devices attached to trees:

• Using technology to restore and revitalize the rainforest

A working paper by the Humboldt Institute for Internet and Society (HIIG) describes how technology plays a vital role in restoring the rainforests effectively. For example, peat swamp forests are particularly effective carbon stores, so maintaining or restoring their health could be something that governments and conservation actors want to prioritize. Drones have been used to disperse seeds, thus contributing to forest (re)growth, but there are concerns about the scalability of this approach. Like the PNAS research team in Peru, the HIIG researchers noted the dependence on incentivized human action on the ground; in other words, technology can only augment what people on the ground are willing to do.

A similar approach has been taken by one of the projects of the FAIR Forward—Artificial Intelligence for All initiative created by GIZ (Gesellschaft für Internationale Zusammenarbeit) and digital.global. In that project, AI technology is used to detect and monitor rainforests in Indonesia that have high carbon content. Gathering such data is the beginning of a better understanding of rainforest areas, of what is happening, and of how human intervention can help to restore damage. For example, fires, like those in Brazil right now, could be detected much earlier.

According to the National Geographic Magazine, drones can help to automate the reforestation process: They map and identify the best places to replant, and then shoot pods into the soil that contain pre-germinated seeds along with the nutrients necessary for healthy growth. According to the authors,  these drones can plant trees in hard-to-reach places at a much lower cost—and up to 10 times faster—than traditional methods.

In all these efforts, complexities of structuring the massive data are very high and identifying  the unexpected and anomalous patterns in vast amounts of data, isn’t simple at all. But the algorithms can learn and reach a high accuracy, as study done by the Rainforest Connection as part of Hitachi’s Social Innovation strategy  shows.

• Making farming more sustainable

The population of our planet is still growing and will continue to do so for a few more decades, reaching 10.3 billion in 2080, according to scientists. We need to break the mantra of “More people born = more food produced = more fertile land required = more forests destroyed.” Above, I discussed how technology can help break this chain at the last step—preventing illegal deforestation—but wouldn’t it be even better if we could increase food production on smaller footprints of land and increase the yield and quality of crops at the same time? This is the big promise of digital farming—often called smart farming, precision farming, or intelligent agriculture.

The idea is that sensors in the soil, metadata from satellites and drones, intelligent agricultural machines, and AI analytics will work seamlessly together to optimize fertilization, irrigation, land use, and machine hours. In doing so, not only will the carbon footprint of crop production be reduced but yields will be increased using less land, as multiple research papers from China, Brazil , and Indonesia have shown.

AI meets jungle—really!

To clarify, AI won’t save the rainforest or create a sustainable world by itself. We still need a global unified agenda, the will of all stakeholders to preserve our natural habitats, and investment in relevant technology. Nevertheless, such technology—and perhaps most importantly, AI—can augment the efforts of communities, activists, governments, NGOs, and—last but not least—companies to create an environment in which we don’t need to destroy our last remaining jungles simply to produce more food and in which all stakeholders can jointly monitor progress and enforce the protection of our remaining rainforests. Many people may think that such a thought is naïve, but I am optimistic that humankind will eventually understand that we have no Planet B and will get its act together to make this happen—for the benefit of everyone and of the generations to come.

(Disclaimer: The ideas, views, and opinions expressed in my LinkedIn posts, articles, videos, and profiles represent my own views and not those of my current or previous employers or any organizations with which I am associated. Additionally, any and all comments on my posts from respondents/commenters to my postings belong to, and only to, the responder posting the comment.)