AI against deforestation: Barry Callebaut and EcoVision Lab unveil large-scale HCS map for Maphilindo

11 May 2021 --- Identifying and measuring deforestation and the cultivation of specific commodities at a large scale and in sufficient detail can be difficult. In line with this, Barry Callebaut has launched a first-of-its-kind, large-scale indicative High Carbon Stock (HCS) map covering Malaysia, Indonesia and the Philippines (Maphilindo).

The company has teamed up with EcoVision Lab, part of the Photogrammetry and Remote Sensing group at ETH Zurich, developing highly automated artificial intelligence (AI) solutions. 

Speaking to FoodIngredientsFirst, Pablo Perversi, chief innovation, sustainability and quality officer, global head of gourmet at Barry Callebaut, says the company is committed to ending deforestation in the cocoa and chocolate industry.

“Working in close partnership and collaboration with industry, governments and civil society, we can deliver on the objectives we set ourselves.”

Identifying at-risk areas
This latest collaboration has led to developing the publicly available HCS map that identifies forests with high conservation value and areas where deforestation would cause the highest carbon emissions.

The group has long-standing experience combining machine learning (deep learning) with remote sensing to address ecological challenges. 

Identifying the link between specific commodities and areas at risk of deforestation can be highly complex. The team at ETH Zurich is utilizing data from a NASA laser scanner attached to the International Space Station and imagery from the European Space Agency (ESA), which allows large areas to be mapped by applying AI, limiting on-the-ground measurements to only very critical locations. 

According to the company, the move marks a “pivotal moment” amid the acceleration of climate change and the extinction of precious flora and fauna. 

A costly exercise
Barry Callebaut requires its suppliers, such as palm and soy suppliers, to identify the forest areas that need protection and those that can be developed for agriculture. 

However, conducting this type of assessment is difficult, time-consuming and often a costly exercise, the company outlines. 

“Best in class” approach
The development of the HCS map supports the current approach of data taken from the field, which is the widely used approach to measure the link between commodity cultivation and deforestation, the so-called High Carbon Stock Approach (HCSA). 

This new method combining deep learning and publicly available satellite imagery is a true breakthrough because it provides a highly automated, transparent, objective tool that generates indicative HCS maps at global scale with unprecedented accuracy,” says Prof. Dr. Jan Dirk Wegner, Head of EcoVision Lab, ETH Zurich and University of Zurich.

HCSA is a widely recognized methodology that is increasingly being used by certification standards, such as the Roundtable on Sustainable Palm Oil (RSPO) and by companies committed to breaking the link between deforestation and land development in either their operations or supply chain. 

The reliance on ground and aerial imagery for HCSA is challenging because manually measuring landscapes and evaluating vegetation classes is labor-intensive and difficult to roll out at scale while using planes equipped with specialized laser scanners is an expensive option. 

Combining HCSA with the power of AI
When implementing AI solutions, it all starts with data quality. Deep learning is a research field with a swift pace.Barry Callebaut has teamed up with EcoVision Lab to develop highly automated AI solutions.

New algorithms are improving quickly and demonstrate the potential to revolutionize forest monitoring and carbon stock estimation based on satellite images. However, when relying on supervised learning, that is, learning from large reference datasets, the amount and quality of data is the key to success. 

Over the past four years, the ETH team has focused on utilizing the new satellite imagery and calibrating regional carbon biomass data. As a result, we have developed a tool that is highly automated, objective and can be used to up-scale indicative HCS mapping to entire world regions.

“AI in combination with a data-driven approach supports our efforts to scale up our impact,” notes Perversi. “We also commenced a large-scale reforestation project to mitigate the impacts of climate change. This project will use a technology-driven approach to pilot the utilization of innovative seedling generation, planting and monitoring in harsh conditions.”

Further commenting on what industry can expect from Barry Callebaut in the years, Perversi says the company will mitigate the impacts of climate change and return the ecosphere of forests. “We will continue to identify the requirements needed for scaling up innovative solutions and investigating the latest opportunities in technology.”

Forever Chocolate
Being carbon and forest positive is one of the key commitments of Forever Chocolate. These maps are the result of a four-year collaboration with one of the leading research institutions combining AI with remote sensing for ecosystem protection. 

“We welcome this innovative new HCS indicative map by ETH Zurich and Barry Callebaut. Mapping the HCS forests in tropical forest regions is a first step toward protecting them from deforestation,” explains Judy Rodrigues, executive director, HCS Approach.

“HCSA is keen to build upon this work from Barry Callebaut and ETH to make high-quality HCS maps available for anyone to use to implement no deforestation, and in particular smallholder farmers.”

Stepping up deforestation efforts
In recent months, there have been several moves made toward protecting forests in cocoa-producing countries. 

Last month, Cargill shone a light on the importance of protecting the forests in Ghana and Côte d’Ivoire. In the company’s latest Cocoa & Forests Initiative report, Cargill revealed that progress and restoration is being made through farm mapping and traceable supply chains.

In the same month, Nestlé stepped up its sustainability efforts to protect and restore forests while also outlining that the iconic KitKat brand will be carbon neutral by 2025.

Outside of cocoa, the palm oil industry is also ramping up its efforts to tackle deforestation. 

Last August, Unilever teamed up with US-based tech company Orbital Insight to boost palm oil sourcing transparency, using geospatial analytics for “sophisticated” deforestation crackdown. 

This approach was touted as bringing a “new level of sophistication to traceability” – one that has the potential to work on a massive scale.

By Elizabeth Green