Praveen Kumar, top right, engages with local field liaisons in an urban area of India to examine the role that AI-powered applications can play in identifying communities that are especially vulnerable to extreme heat. Courtesy photo.
“Can we use artificial intelligence to help people sustainably adopt new technologies, improve their lives, and reduce their vulnerability to climate and environmental risks?”
That is the primary question driving the research of Praveen Kumar, an associate professor at the Boston College School of Social Work who is exploring how AI can help vulnerable communities in India prepare for climate-related threats.
Kumar and his research team are pursuing answers on two fronts in India. In the rural and tribal communities in Odisha and Jharkhand, they’re studying how AI-powered digital tools can help people understand climate risks and make locally informed decisions about how to adapt. In the urban areas of Karnataka and Tamil Nadu, they’re developing AI-powered early warning systems designed to help communities prepare for extreme heat.
In Odisha and Jharkhand, Kumar and his colleagues are working with communities whose livelihoods depend on shared resources such as water, pastures, and farmlands. The researchers are studying whether a suite of digital applications—Gramify, Samaaja, Know Your Landscape, and Commons Connect—can move beyond the delivery of static weather reports to help residents document changing conditions and participate in decisions about managing those resources.
Kumar’s team is using a decision-making framework called fuzzy cognitive mapping to ask communities how they understand climate risks, what information they need, and what they want the apps to provide. The approach, he said, enables residents to identify how these tools could better support their decisions.
“FCM enables us to create a blueprint for a revised application,” said Kumar, whose research focuses on helping vulnerable communities adapt to climate change. “Our goal is to ensure that the algorithmic backend of these AI-powered tools actually aligns with local logic, treating communities not as passive data points, but as active co-designers of their own adaptation interventions.”
Praveen Kumar.
His other project tackles another urgent threat: extreme heat in Karnataka and Tamil Nadu. In partnership with the M.S. Swaminathan Research Foundation and the Indian Institute of Science, he is using AI to build hyper-local risk models that combine real-time meteorological data with health and social indicators to identify communities that are especially vulnerable to extreme heat. Instead of just predicting the weather, the AI models the vulnerability of specific neighborhoods.
Kumar’s team plans to share heat warnings with local residents through mobile phone networks while training health officials to better prepare for extreme heat emergencies.
He said working directly with communities is essential to making AI useful. It’s one thing to develop an AI-powered tool; it’s another to help communities see its value.
“Technology is only as good as the social infrastructure supporting it,” said Kumar. “If a predictive model issues a local heat alert, or a global review identifies a severe regional data gap, but community members don’t trust the platform or lack the institutional capacity to respond, the technology adoption is bound to fail.”
This is where social workers come in. By treating communities facing climate risks as vital sources of lived experience, local knowledge, and collective power, social workers can help bridge the gap between AI developers and the people they hope to serve.
“I want all our social work researchers and practitioners to leverage AI to do good, to improve its productivity, and serve humanity,” said Kumar. “I know that with the competencies the social workers have, they absolutely have a seat at this table.”
This story is part of our series highlighting how students, faculty, and staff at BCSSW are putting AI into practice across teaching, research, and clinical training.
