Machine Learning Can Help Track Climate Technology Innovation
tracking the development and scaling of climate technologies remains a challenge. machine learning can help
We use large language models and LinkedIn data to construct a global network of key public and private organizations collaborating on climate-tech innovation. The resulting network encompasses 134,727 organizations across 189 countries. It includes diverse types of collaborations, such as R&D partnerships and equity investments, and spans 19 climate technologies.
Driving Questions
How can machine learning and large language models enhance our ability to track and analyze global climate-tech innovation and collaboration?
What role do governmental organizations play in fostering climate-tech innovation clusters, and how can policy interventions better support technological diversification and commercialization in emerging regions?
our methods
big data + Natural Language PRocessing
We use a novel machine learning approach to track climate-tech innovation worldwide. By analyzing over 20 million LinkedIn posts from key public and private organizations, we’ve built a comprehensive network of climate-tech collaborations, including R&D, demonstration projects, and commercialization activities. Each collaboration is classified by the specific climate technology, type of partnership, and the organizations involved.
We also categorize organizations by type and sector, allowing a deeper understanding of each actor's role in climate innovation. Our method begins with an initial set of climate-tech organizations, then iteratively expands by adding new ones from identified collaborations. This approach offers a detailed, real-time view of global climate-tech innovation networks.
Geographic distribution of organizations engaging in climate-tech as of April 2024
major findings
Machine Learning Can Help Tracking Climate-Tech Innovation
Concentration of Innovation in Limited Technologies: Around 60% of new climate-tech startups and activities are concentrated in just three technologies—solar, electric vehicles, and hydrogen—raising concerns over insufficient innovation in other critical areas like heat pumps, biofuels, and carbon capture, which are also vital for meeting global climate targets.
Central Role of Governmental Organizations: Governmental organizations are pivotal in forming and supporting climate-tech innovation clusters, especially for commercialization. This influence is notable even in regions with less established industry presence, highlighting the potential for government-led innovation hubs.
Need for Technological Diversification: The study suggests that to meet climate goals, governmental organizations must foster greater technological diversification by supporting innovation in a broader range of climate technologies across different regions.
