AI-powered renewable planning cuts underserved by 96% in US cities
Receding-horizon MDP policy hits 66% green penetration while slashing inequity
A new arXiv study tackles the challenge of equitable renewable energy planning by framing budget allocation as a Markov Decision Process (MDP). The research, led by Riya Kinnarkar and colleagues, uses a single solver interface to compare two distinct governance models: mature economies where government directly builds generation (tested on eight US cities) and emerging economies where planners steer private investment through incentives and quotas (West Java, Indonesia).
Results show a receding-horizon value-iteration policy dominates in both settings. In the US, it drives 66% renewable penetration while cutting the underserved low-income population by 96% versus a random baseline. In West Java, it closes the low-access gap while crowding in the most private capital. However, a naive market-chasing heuristic—only mildly suboptimal in the US—causes catastrophic outcomes in Indonesia by abandoning every low-access region, because chasing attractive markets and serving equity goals diverge when planners act through private developers. The study highlights that policy success depends heavily on the economic context and governance structure.
- 66% renewable penetration achieved in US cities with 96% reduction in underserved low-income population vs random baseline
- In West Java, the same policy closes the low-access gap while maximizing private capital investment
- Market-chasing heuristic is mildly suboptimal in mature economies but catastrophic in emerging ones, underserving all low-access regions
Why It Matters
Framework helps energy planners balance equity and efficiency across vastly different economic contexts using AI-driven policy optimization.