Using artificial intelligence to help restore the natural world

There is money there to help the rainforest; but there has to be proof of care. Artificial intelligence, as captured by GainForest, can help.

It is not a fair fight. Indigenous residents, ancestors stretching back generations, are struggling to keep their forests intact. Against them: loggers, industrial farmers, government agencies, hunters.

The implications are not just local. They are global. Climate change and biodiversity loss are two bullets in the same gun. Perils we face in parallel, both driven by deforestation, and at a rate unseen in human history. If global deforestation were a country, its emissions would be larger than the whole of the European Union.

The dangers are recognised. Funds are needed. To preserve their forest reserve in the Amazon, the Brazilian Kayapo people require USD1.2 million a year, a huge amount for an economy largely based around subsistence farming. There is money on the other side. Around USD25 billion has been set aside to preserve global forestry. However, it can only be distributed against results. Overlapping land claims and slow performance measurement make funds inaccessible.

We cannot value what we cannot measure.
David Dao, AI for Sustainable Development, ETH Zurich

Standard measures of forestry are slow, costly and not always accurate. Manual monitoring and verification of forest can cost up to USD300 per hectare. Satellite imagery reduces time, but the low resolution pictures are not fully transferable to what might exist on the ground.

The two sides have to be brought together. We cannot value what we cannot measure. Enter, from the other side of the world, GainForest. GainForest collaborates with a combination of local inhabitants, high resolution drone imagery and low resolution satellite imagery.

Working with agencies to verify the compatibility and reliability of the different data sources, GainForest is committed to transparency in its processes. Data sources are agglomerated and fed in a learning loop into an AI predictor. The analytics of the AI programmes not only estimate the level of carbon captured in the forest, it also provides data on tree tracking and species classification.

Proof of care can be converted into digital assets.
David Nao, AI for Sustainable Development, ETH Zurich

When measured in samples against on-the-ground findings, the AI predictions have proved remarkably accurate. Trust in AI is growing, and GainForest plans to show its technology to audiences at the 2021 Glasgow COP26 Climate Change Conference.

Measurement has to be translated into value. Machine learning provides the proof of care. Stakeholders accept the evidence and unlock the funding. Proof of care can be converted into digital assets. The carbon value is converted into cryptocurrencies and distributed on blockchain technologies, thus cutting out the middleman. NGOs stand at the penultimate link of the chain, finally redistributing tokens to the indigenous people acting as guardians of the forest.

GainForest currently runs four forestry protection programmes. It is a model that can be run to greater scale; and proof that artificial intelligence can play a valuable role in our efforts to stabilise our climate and avert deforestation.

Summary based on Swiss Re Institute event: Algorithms for hope.

Algorithms for hope

Case studies with a positive global impact as we move from human to augmented intelligence. Swiss Re Institute in partnership with Gottlieb Duttweiler Institute and IBM Research.