Machine Learning Maps How Ready Small Firms Are for the Net Zero Transition

Micro, small, and medium enterprises are the quiet giants of the global climate problem. Individually they may seem too small to matter, but collectively they account for a substantial share of worldwide greenhouse gas emissions, and in India they form the backbone of industrial employment and output. A new study published in Discover Sustainability by researchers at IIT (BHU) Varanasi, Symbiosis International University, and GLA University argues that the missing piece in India’s net zero puzzle is not ambition but assessment: before anyone can design a decarbonization strategy for these firms, someone has to measure how prepared they actually are. The team, led by Gaurav Upadhyay and Cherian Samuel of the Department of Mechanical Engineering at IIT (BHU), together with Dilawar Ahmad Bhat and Nitesh Thakur, has built a machine learning assisted maturity model that sorts Indian MSMEs into four levels of net zero readiness, and the results are sobering.

The central finding is stark. When the model was applied to a varied cohort of MSMEs operating across different domains in India, including firms in Eastern Uttar Pradesh, only 8 percent were operating at a consolidated maturity level, the highest of the four. Seventeen percent sat at an evolving level, 49 percent at a primary level, and 26 percent at an insufficient level, the lowest rung on the ladder. Put differently, three quarters of the assessed firms occupy the earliest two maturity stages, meaning they are either barely aware of net zero practices or have just begun to engage with them. For a sector that policymakers hope will drive both economic growth and emissions reductions, that distribution reads as a warning label.

What makes the study technically interesting is how the maturity levels are assigned. Rather than relying on a single questionnaire score, the researchers decomposed organizational readiness into four parameters: awareness and engagement, knowledge and capabilities, planning and strategy, and physical actions. Each parameter captures a different dimension of the transition. Awareness and engagement measure whether a firm’s leadership and workforce even recognize net zero as a relevant concern. Knowledge and capabilities probe whether the firm possesses the technical understanding and skills to act on that recognition. Planning and strategy assess whether readiness has been translated into formal goals and roadmaps, while physical actions capture the ground truth of what has actually been implemented on the factory floor. Every attribute in the model is mapped to supporting literature or expert rationale, a transparency step documented in the study’s appendix.

Data came from surveys and interviews with MSMEs and domain experts, gathered under ethical approval from the Institutional Ethics Committee of the Department of Mechanical Engineering at IIT (BHU), with informed consent obtained from all participants in line with the Declaration of Helsinki. The researchers then applied data analytics techniques and supervised machine learning classification algorithms to categorize each firm into one of the four maturity levels. Supervised classification is the branch of machine learning in which an algorithm learns from labeled examples, firms whose maturity level is already known, and generalizes to assign labels to new cases. This approach allows the model to capture nonlinear relationships between survey responses and maturity outcomes that a simple weighted checklist would miss, and it makes the assessment repeatable as more firms are added to the training data.

The choice of a maturity model as the underlying framework is deliberate. Maturity models, long used in software engineering and quality management, describe organizational capability as a staged progression from ad hoc, unstructured practice toward optimized, institutionalized behavior. Applying that logic to net zero readiness lets a firm locate itself precisely on a developmental path rather than receiving a binary pass or fail verdict. A firm at the insufficient level needs fundamentally different interventions, basic awareness campaigns and leadership engagement, than a firm at the evolving level, which may need technical training, energy audits, or capital access to move from planning to physical action. The staged structure therefore converts a diagnostic score into a prioritized action plan.

The authors position their contribution against the existing research landscape, noting that prior studies have primarily focused on individual sustainability dimensions, such as energy efficiency or waste management in isolation. The proposed framework instead provides a comprehensive net zero maturity assessment tailored to the Indian MSME context, where customized strategies have been notably scarce. This matters because MSMEs differ from large corporations in crucial ways: they operate with thin margins, limited technical staff, and little slack for compliance overhead. A maturity framework designed for multinational supply chains cannot simply be scaled down, and the study’s four parameter structure reflects the organizational realities of small firms rather than the reporting apparatus of large ones.

The regional framing of the study, centered on Eastern Uttar Pradesh, adds a layer of significance for India’s industrial geography. The region is home to clusters of small scale manufacturing, from metalworking and textiles to food processing, where enterprises often lack direct access to the sustainability consulting and green finance channels available in metropolitan centers. The corresponding author, Dilawar Ahmad Bhat of the Symbiosis School of Banking and Finance, along with co author Nitesh Thakur of GLA University’s Institute of Business Management, brings a management and finance perspective that complements the engineering expertise at Varanasi. The interdisciplinary composition of the team is visible in the model itself, which treats readiness as simultaneously a technical, organizational, and strategic phenomenon.

For practitioners, the practical payoff is gap identification. By scoring themselves against the four parameters, MSMEs can see exactly where their organizational readiness is weakest and prioritize actions that support the net zero transition in a sequence that matches their maturity stage. A firm strong on awareness but weak on physical actions knows its bottleneck is capability or capital, not motivation. A firm weak on awareness knows that even the best available technology will sit unused without leadership engagement. The machine learning layer strengthens this diagnostic by learning from the accumulated experience of the surveyed cohort, allowing newly assessed firms to benefit from patterns detected across hundreds of responses rather than from a static rubric frozen at publication.

The study received no specific funding from any institution, organization, or funding agency, and the authors declare no competing interests. It was received by the journal on 19 May 2026, accepted on 21 September 2026, and published on 16 October 2026 as an open access article under a Creative Commons Attribution NonCommercial NoDerivatives license, making the full model and its supporting material freely available to researchers, policymakers, and enterprise support organizations. The work is citable under its permanent DOI, 10.1007/s43621-026-04826-x, and the published version carries the standard peer review guarantees of the journal.

The broader implication reaches beyond India. Global warming and climate change have emerged as the predominant issues of this decade, and MSMEs substantially contribute to global greenhouse gas emissions wherever they operate. If three quarters of assessed firms in one of the world’s largest MSME economies sit at the earliest two maturity levels, the same pattern likely holds in many developing and emerging markets. The study suggests a replicable template: define readiness parameters grounded in literature and expert rationale, collect structured data from firms and experts, train supervised classifiers to assign maturity levels, and use the resulting distribution to target policy interventions where they will move the most firms up the ladder. As net zero deadlines approach, the ability to measure preparedness at scale may prove as consequential as the technologies of decarbonization themselves.

Subject of Research: Sustainability

Article Title: A machine learning assisted net zero maturity model for MSMEs in Eastern Uttar Pradesh India

Article References: Upadhyay, G., Samuel, C., Bhat, D. A., & Thakur, N. (2026). A machine learning assisted net zero maturity model for MSMEs in Eastern Uttar Pradesh India. Discover Sustainability. https://doi.org/10.1007/s43621-026-04826-x

Image Credits: AI Generated

DOI: 10.1007/s43621-026-04826-x

Keywords: Net zero, MSME, Machine learning, Maturity model, Climate change, India, machine, learning, assisted, zero, maturity, model