What problem were they solving?

Traditional brick kilns harm the environment, through their high usage of fertile topsoil to make the bricks and the nature of emissions from the chimneys during the firing process.  Brick manufacturing contributes eight percent of the air pollution in Delhi and its surrounding districts. In addition, workers at brick kilns often face forced labor conditions.

What did they do?

The UNDP Accelerator Lab in India and University of Nottingham developed a new methodology using artificial intelligence – combining machine learning algorithms and geospatial analytics – to map the entire brick kiln belt in India, which has been used by the Bihar State Pollution Control Board to better target environmental policy violations. The GeoAI digital platform detects hotspots of air pollution using satellite imagery and computer vision algorithms (in this case the same algorithm used to identify dog breeds). The partners worked with citizen scientists on the Zooniverse platform to create a labeled dataset of satellite images. This was used to train a computer vision algorithm to detect the specific brick kilns which are hotspots of vulnerable labor and air pollution, uncovering non-compliance with environmental policy in India.

What was the benefit of using collective intelligence for this issue?

As a result, more than 47,000 brick kilns have been detected across Indo-gangetic plains of India and incorporated by UNDP into the GeoAI open data platform. The GeoAI platform is also used to crowdsource reports of violations of labor laws, human rights and social security regulations. Using GeoAI in the State of Bihar, the total number of brick kilns was brought down to a manageable number for staff to inspect. Around 7,500 brick kilns were first analyzed by GeoAI, and it was determined that 1,655 kilns were high risk. Environmental regulators were then able to complete an inspection of 1,013 of those, which led to the green transition of 1,000 brick kilns and the reduction of 500,000 tons of CO2 per annum (equivalent of 100,000 gasoline vehicles). Loans are being secured to help families transition to greener livelihood alternatives.

What does this experience tell us about collective intelligence for climate action?

This experiment brings together coordinated action from diverse stakeholders – regulators, government agencies, civil society and volunteer groups to tackle inaction on environmental commitments. The platform demonstrates the value of automated approaches for improving the efficiency and scale of compliance monitoring efforts. UNDP India is now scaling out this platform to two additional Indian states and in Nepal.