How AI-powered satellite data could boost the organic cotton industry

A new programme led by the Global Organic Textile Standard would use satellite and AI technology to detect fraud and monitor growth across farms growing organic cotton.
How AIpowered satellite data could boost the organic cotton industry
Photo: GOTS

To become a Vogue Business Member and receive the Sustainability Edit newsletter, click here.

The Global Organic Textile Standard (GOTS), in partnership with the European Space Agency and AI company Marple, is using satellite data to detect organic cotton fields across India. It’s an effort to strengthen the integrity of organic certification, provide more accurate organic yield forecasts and open up pathways for more farmers to become certified. The ultimate goal, say the organisers, is to increase the total supply of organic cotton and benefit small farmers in the process.

The project, announced today, plans to use remote monitoring technology that involves training AI to use satellite data from the European Space Agency to detect and then classify cotton fields according to their cultivation standard. GOTS can then distinguish between fields using organic and nonorganic cultivation methods. It’s not unlike facial recognition technology used to unlock smartphones, says GOTS project manager Jeffrey Thimm — “but instead of facial features, we are looking at parameters like growth rate and vibrancy over a growing season”.

Fashion brands have set goals to source more organic cotton in the next five to 10 years, recognising the resource-intensive agriculture practices and chemicals associated with traditional cotton production. But organic cotton has a supply problem: as of 2020, it accounted for less than 1 per cent of all cotton production, down from a decade earlier. If successful, the project could help to mitigate two of the biggest challenges facing the organic fibre sector: concerns about fraud within the supply chain, as well as the limited supply.

A satellite image of cotton fields in Uzbekistan remotely detected in 2022 using Marple
s technology.

A satellite image of cotton fields in Uzbekistan remotely detected in 2022 using Marple's technology.

Photo: Marple, with modified Copernicus Sentinel data processed by Sentinel Hub

Fields that are certified organic but classified by the technology as non-organic can be flagged for investigation by the responsible certifying body; that could help to prevent fraud before the cotton is harvested and enters the supply chain. Thimm expects the project to boost confidence in the integrity of organic cotton supply chains through other means, such as by identifying risk of contamination from genetically modified (GMO) cotton. “Knowing if there are GMO cotton fields near a community of organic cotton farmers will help in risk assessment and mitigation for preventing inadvertent contamination,” he says.

At the same time, many farmers are believed to practise organic agriculture even though they are not certified as such — meaning they do not use synthetic chemicals or fertilisers, for instance, but have not gone through the costly and time-consuming paperwork or verification process involved in being able to use the organic logo or earn the price premiums associated with organic cultivation. The use of remote sensing to detect farms that may fall into that category, the project planners say, could help those farmers to become certified. Thimm says one of GOTS’s underlying goals is to reduce the barriers to sustainability and increase participation of marginalised farmers.

“If we can remotely identify so-called ‘uncertified organic’ cotton fields, these farmers can be easily invited to join a farm-group project that supports the process of conversion to certified organic. Furthermore, by supporting the expansion and training organic farm groups, the cost of certification can be greatly reduced.”

They are aiming to differentiate between organic; in-conversion to organic (years one to three of using organic practices); natural (uncertified or almost organic); high-input non-GMO; and GMO. They don’t know yet if they can differentiate between all five of those categories, but the list “approximately maps the spectrum of environmental impact (low to high, respectively) based on chemical inputs”, says Thimm.

Co-financed by GOTS and the European Space Agency, in collaboration with German software development firm Marple GmbH, the project builds on a feasibility pilot that was done in Uzbekistan in 2021, and will be deployed in the three distinct cotton-growing regions of India. Potential for future expansion includes other major cotton-producing regions such as Türkiye and parts of Africa, as well as use in other fibre crops. 

Comments, questions or feedback? Email us at feedback@voguebusiness.com.

More from this author: 

Getting closer to measuring fashion’s true carbon footprint

EU proposal must hold fashion brands accountable for global textile waste, advocates say

Viscose alternative Spinnova has a wood pulp problem