Using digital technologies to protect, restore and promote life on land

United Nations Sustainable Development Goal 15 seeks to “protect, restore and promote sustainable use of terrestrial ecosystems, sustainably manage forests, combat desertification, and halt and reverse land degradation and halt biodiversity loss” (United Nations, n.d.).

Deforestation – the practice of cutting down or burning large areas of forests – has robbed our natural world of 420 hectares of woodland in the last two decades – that is 27 soccer fields of trees every minute for twenty years (earth.org). The importance of forests “cannot be underestimated” according to the World Wildlife Fund (2020), which stresses forests’ importance for the air we breathe, landscapes around us, other plant and animal life, the stable climate, and as the basis of the employment upon which over 20% of the world’s people rely.

Yet we continue to deforest huge areas of land to make way for products in high demand in developed economies, such as beef, cocoa, coffee, and vegetable oils (earth.org).

Introducing artificial intelligence

Artificial intelligence (AI) is a field of study focused on work which can be done entirely by machines – which learn and adapt independently of human input – but which previously had needed human facilitation. Initially this conceptualization focused on machines’ potential ability to use language and to solve problems (McCarthy et al., 1955). AI has subsequently been developed to solve problems and do human tasks, such as playing chess, better than humans can (Lowry, 2021). Using large datasets, algorithms are produced to both improve and accelerate tasks previously done by humans (Pournader et al., 2021). Interest in AI has accelerated rapidly in the last decade as its potential for generating value has developed. In a forestry context, for example, satellite images of areas where deforestation is suspected can be fed into an AI system, which could identify where forest boundaries are changing (Leal et al., 2020).

The challenge for Unilever

Unilever is one of the largest buyers of palm oil, for use in its food, cosmetics and hygiene products. Driven by European Union sustainability legislation, environmental and indigenous rights campaigns and consumer concern, Unilever has been attempting to create a more sustainable palm oil industry. Unilever has recently made a commitment to achieve a deforestation-free supply chain by the end of 2023. Most companies sourcing palm oil find it difficult to assess the actions of growers in their supply chain, particularly in terms of deforestation.

Unilever has backed a digital initiative that bring together artificial intelligence, satellites, geospatial data and blockchain technologies to improve its sourcing. The focus is to put the spotlight on the first mile where palm oil is grown, collected, transported and processed (Bryce 2022). The palm oil sector must address social, environmental and economic issues to be truly sustainable.

How artificial intelligence is being used

Unilever is working with several firms to improve transparency in the first mile of palm oil supply chain. Unilever works with the market research firm, Premise, to create a community of local people who track the collection of palm oil. Mill workers and suppliers provide photos and information about collection points via Premise’s digital crowdsourcing platform. The platform analyses materials that are input in a quality control system, powered by AI, to verify their credibility and ensure collection points aren’t double-counted. In North Sumatra and Aceh provinces, 5000 collection points have been identified and documented.

In 2021, a new partnership with Meridia Land was formed to facilitate digital mapping of the smallholders in areas supplying the mills. To date, the mapping includes over 9,000 smallholder land parcels and aims to map over 30,000 smallholders in and around sourcing hubs. Analysis of satellite imagery allows the identification of any deforestation that has taken place historically or recently at these suppliers. Unilever can then incorporate this information into their digital ecosystem to identify the sourcing relationships and their historical and ongoing practices to ensure that their supply chain meets their sustainability objectives.

In addition to employing digital technologies in the supply chain, Unilever has introduced a range of other measures to reduce risks associated with deforestation such as concentrating the sourcing of palm oil from areas with lower risk of deforestation and partnering with suppliers that share Unilever’s sustainability ambitions. There will also be a reduction in the number of palm oil mills in their network from 1600 to around 500.

Beyond palm oil

The approach taken by Unilever can also be used for other crops associated with deforestation, whether it’s soybeans in Brazil, rubber in Malaysia, or paper pulp in Thailand. The geospatial and AI-driven technology supports continuous monitoring of planted and natural forests, grasslands and agriculture, water resources, buildings and roads.

Lessons learned for other organisations

1. There are newer digital technologies that can be brought together in the supply chain that automate the detection of issues such as deforestation.
2. Monitoring and detection in the supply chain typically need to be part of a wider strategy that is shared by supply chain partners, as well as worker and indigenous organisations.
3. Large multi-national companies have a significant amount of influence in developing countries and accordingly a huge responsibility to act ethically and be socially responsible. However, even smaller companies can have influence if they coordinate with other companies and organisations.