CS3: Innovation, R&D and Technology Policy for the Transition

Public subsidy schemes for residential energy retrofits are often designed as uniform financial instruments. Yet retrofit decisions require households to evaluate upfront costs, expected energy savings, and the effort involved in applying for support. These decision processes may differ across households depending on financial understanding and energy-related knowledge.
This paper examines how such differences shape responses to energy efficiency incentives in the Netherlands. The analysis is based on a discrete choice experiment conducted among 796 Dutch homeowners. Mixed logit models are used to estimate preferences for investment costs, projected savings, subsidy levels, timing of incentives, and application support. The models allow responses to vary with investment literacy and energy-related attitudes and habits.
The results show clear heterogeneity. Financially literate homeowners are more likely to invest and more responsive to projected savings and incentive levels. By contrast, households with lower energy literacy attach greater value to assistance in the application process. Changing the timing of financial incentives alone does not substantially alter participation among lower-income or less financially literate households, whereas procedural support increases adoption probabilities more broadly.
These findings suggest that the effectiveness of energy efficiency policy in the Netherlands depends not only on the level of financial support but also on how accessible and administratively manageable the schemes are. Accounting for differences in financial and energy literacy can help improve policy design without necessarily increasing subsidy levels.

Despite huge efforts to integrate renewable power into the energy sector, the reliance on fossil fuels remains a reality for the foreseeable future, especially in hard-to-abate sectors such as steel, cement, and chemicals where electrification is not always feasible. The associated emission of carbon dioxide, therefore, has proven to be a critical challenge for the world. Thus, the accelerated development of green technologies to tackle carbon accumulation in the atmosphere has been of immense interest globally. That is evident in the fast-adapting development methodologies for carbon capture and utilization (CCU) and green fuel technologies from traditional modeling to the integration of artificial intelligence and machine learning for collection and analysis of larger amounts of data, and now to adaptive learning and closed-loop autonomous laboratories for high-throughput tangible results. 



The very large number of possible materials, and their combination with the space of reaction conditions presents an opportunity to capitalize on fully automated or limited manual labor labs to accelerate the design and discovery of the optimum reaction setup. The existing literature reviews delve into machine learning applications in carbon capture and carbon conversion processes, but adaptive learning and closed-loop autonomous labs are still unexplored.



This paper aims to review recent advancements, current challenges, and the huge potential of autonomous and limited manual labor labs. It will also explore their role in fast-tracking the development of CCU and green fuels' technologies through combining Bayesian optimization and automated systems. Combining artificial intelligence and robotics can further advance and accelerate material discovery and process optimization. The review explores this cutting-edge approach with focus on carbon capture and conversion to valuable green fuels and chemicals to serve as a valuable resource for research targeting the development and optimization of applicable CCU systems and tackle carbon emissions globally. 

The climate crisis urges a profound structural transformation of the global energy system, shifting from carbon-intensive energy sources towards renewable energy sources. However, despite their undeniable benefits in reducing GHG emissions, the transition to renewables poses substantial challenges. An energy mix dominated by renewables can threaten national energy security and system stability due to their volatility and intermittency. In this context, AI emerges as a key technology to facilitate the integration of renewables into the energy mix by reducing the mismatch between energy supply and demand through its advanced predictive capabilities and accurate forecasting of climatic and weather conditions. This study conducts an analysis of the impact of AI on the low-carbon energy transition (LCET) from both the supply and demand sides across OECD countries, addressing the question of whether AI can be an effective tool for boosting LCET. 
The study identifies the channels through which AI influences LCET on both the demand and supply sides using panel mediation analysis. In addition, it examines heterogeneous effects of AI on LCET by using a K-means clustering. Finally, difference-in-differences (DiD) is conducted to see whether climate policy stimulates the sustainable use of AI. 
The findings indicate that, on the supply side, AI facilitates LCET by supporting renewable energy deployment, improving energy efficiency, and driving economic restructuring toward less energy- and pollution-intensive sectors. On the demand side, AI affects LCET by stimulating technological innovation and encouraging sustainable consumption behaviours among citizens. Moreover, the impact of AI on LCET is heterogeneous across OECD countries, and climate policy may induce the sustainable use of AI, thereby further accelerating LCET.
Given the crucial role of OECD in achieving net-zero goals, our study provides useful guidelines on how countries can effectively harness AI to accelerate the LCET.

This paper examines whether mandatory regulatory obligations can operate as effective innovation policy instruments in capital-intensive energy sectors. It analyzes the Research, Development, and Innovation (RD&I) Clause embedded in Brazilian oil and gas exploration and production contracts since 1999. The research question is: to what extent has the RD&I Clause contributed to technological capability building, scientific infrastructure development, supplier upgrading, regional diversification, and alignment with energy transition objectives over the past twenty-five years?
The study adopts a qualitative and documentary approach combining regulatory analysis of the Clause's evolution under Law No. 9,478/1997 and subsequent ANP regulations-particularly Resolution No. 918/2023-with systematic examination of publicly available data on investments and projects between 1999 and 2025. The empirical assessment covers more than R$ 37.8 billion in mandatory investments and over 15,000 projects. The analysis considers allocation by technological area (exploration and production, cross-cutting technologies, downstream, and low-carbon initiatives), technological maturity level (from applied research to pilot units), executing agents, and regional distribution.
Findings indicate that the RD&I Clause has generated a stable, revenue-linked funding mechanism for long-term innovation. Investment patterns evolved from an initial emphasis on human capital formation and laboratory infrastructure toward higher technological readiness levels, including experimental development and pre-commercial validation. The instrument has strengthened collaboration among operators, universities, research institutes, suppliers, and startups, while progressively incorporating projects related to decarbonization, energy transition, energy efficiency, and digital transformation.
Policy improvement opportunities include reinforcing strategic governance with clearer long-term priorities, adopting impact-oriented monitoring beyond financial compliance, updating regulatory definitions to better support technology transfer and scale-up, and expanding the effective participation of domestic suppliers and technology-based firms. The paper concludes that RD&I obligations can function as adaptive regulatory instruments that foster technological upgrading, support energy transition goals, and enhance industrial competitiveness.