Research Worth Reading

Materealistic? How European energy system models exceed raw material reserves — Systematic review of 59 highly decarbonised European energy system scenarios showing that most models neglect critical raw material demands, raising questions about material feasibility. Highlights a critical gap in energy transition planning that engineers and modelers need to address.

An Admittance-Based Inverter Connection Screening Tool for Small-Signal System Strength — Presents a screening tool for assessing the impact of new inverter-based resource (IBR) connections on small-signal stability, addressing sub-synchronous oscillations in high-IBR-penetration power systems. Directly relevant to grid planning engineers working on renewable integration.

AI4Land: Scalable Deep Learning for Global High-Resolution Land Use Reconstruction — Introduces a data-driven framework for generating high-resolution historical and future land use reconstructions to reduce uncertainty in terrestrial carbon cycle modeling. Addresses a key limitation in Earth system models used for climate projections.

Violation-Informed Spatio-Temporal Adaptive Targeting Framework for EV-Driven Distribution System Expansion Planning — Proposes a computationally efficient STAT framework for distribution network expansion planning that accounts for EV-induced voltage drops and line overloads. Addresses a critical scalability challenge for utilities facing rapid EV adoption.

Technology & Innovation

This AI Software Saved New Yorkers $5 Million in Heating Costs — EnTech’s Series S2 AI-driven heating optimization software saved New York City residents $5 million in heating costs by reducing energy waste in buildings, which account for ~70% of the city’s emissions. Concrete case study of AI applied to building energy efficiency at scale.

GM Activates Vehicle-to-Grid (V2G) for Existing Customers, Expands Sodium-Ion Grid Storage, and Launches Energy Pass — GM announced three major energy initiatives: enabling Vehicle-to-Grid (V2G) functionality for existing EV customers without requiring new hardware, scaling up grid-scale sodium-ion battery storage, and introducing ‘Energy Pass’ as a universal interface for energy management. Signals a shift toward bidirectional EV integration and alternative battery chemistries for stationary storage.

Open Source Projects

America’s Biggest Solar Factory Is Nearly Complete — QCells is nearing completion of the largest solar factory in the United States, representing a major milestone in domestic solar manufacturing capacity. Significant implications for the U.S. clean energy supply chain and solar deployment scalability.

Today’s Synthesis

The convergence of GM’s V2G rollout, the Violation-Informed STAT framework for EV-driven distribution planning, and EnTech’s building heating optimization signals a maturing software layer for distributed energy resource (DER) orchestration. GM’s hardware-free V2G activation means millions of EVs can soon act as grid assets; the STAT framework gives utilities a computationally tractable way to plan distribution upgrades around EV-induced voltage and thermal violations; EnTech proves AI-driven load shifting delivers measurable savings at city scale. For a software or ML engineer, the actionable opportunity is building the control plane that sits between these layers: grid-aware scheduling algorithms that co-optimize EV charging/discharging, building thermal loads, and stationary storage against real-time distribution constraints (voltage, line limits, transformer aging). This requires differentiable power flow solvers, stochastic optimization for forecast uncertainty, and APIs that aggregate heterogeneous DERs into virtual power plants — problems squarely in the wheelhouse of engineers who’ve built real-time control systems, fleet orchestration, or large-scale simulation pipelines. The grid is becoming a software-defined network; the missing piece is the orchestration stack.