Terra Daily — May 12, 2026

Research Worth Reading

• Forecasting Residential Heating and Electricity Demand with Scalable, High-Resolution, Open-Source Models — An open-source probabilistic deep learning framework for high-resolution residential heating and non-heating electricity demand forecasting, trained on real consumption data. Directly applicable to grid load forecasting, demand-side management, and energy system modeling — useful reference material if you’re building load prediction pipelines.
• Price Distortions in Korea’s Electricity Market: Barriers to Renewable Integration and Reform Pathways — Analyzes structural price signal distortions in Korea’s cost-based pool electricity market that undermine renewable energy integration. Proposes reform pathways to better reflect transmission constraints, real-time supply/demand dynamics, and generator-specific costs — relevant if you’re interested in wholesale market design or grid optimization.
• A Fast-Path to Affordability: Understanding the Benefits of Energy-Only Resources in PJM — Examines how PJM can accelerate grid interconnection for energy-only resources (generation without capacity payments) to reduce costs and deploy new capacity faster. Offers concrete insights into market mechanics and renewable energy integration timelines.
• METBRA25Y: Brazil Surface Meteorology Archive with Harmonized Variables and Quality Control — A harmonized archive of 25+ years of hourly surface meteorological observations from Brazil, derived from INMET public records with quality control applied. Useful as open training or validation data for climate, hydrological, agricultural, or ML studies in the region.
• From Expansion to Consolidation: Socio-Spatial Contagion Dynamics in Off-Grid PV Adoption — Studies how social ties embedded in physical space drive residential solar PV adoption in off-grid rural regions, providing empirical evidence on diffusion dynamics. Relevant for engineers and planners designing off-grid solar deployment strategies or modeling adoption curves.

Technology & Innovation

• Introducing Ford Energy — Ford formally launches Ford Energy as a dedicated business unit after a year of securing supply chains and aligning manufacturing. Signals a major OEM’s strategic bet on energy services — V2G, home energy storage, and grid-interactive vehicles — which opens up roles for embedded systems and controls engineers.

Open Source Projects

• 370 Tesla Semi Trucks Ordered — WattEV orders 370 Tesla Semis, with 300+ headed for the Port of Oakland under a joint program. One of the largest fleet orders to date; useful case study for anyone working on heavy-duty EV logistics, charging infrastructure sizing, or port electrification planning.
• Software is helping this real estate giant burn less gas in NYC — AvalonBay is deploying Parity’s building energy management software across NYC properties to cut gas consumption under Local Law 97’s net-zero-by-2050 mandate. A real-world example of building decarbonization software operating at commercial portfolio scale.
• Cross Country Road Trip in an Old Tesla Model 3 — Documents a 7-year-old Model 3 with 175k+ miles, tracking battery degradation against its original 310-mile EPA range and testing charging infrastructure across a cross-country route. Practical long-term EV battery performance data for engineers evaluating degradation models or fleet lifecycle assumptions.

Community Finds

• The Ocean Is Not A Server Rack: Panthalassa, Peter Thiel, And Wave-Powered AI Compute — A critical engineering analysis of Panthalassa’s proposal to power AI data centers with ocean wave energy. Scrutinizes the feasibility of autonomous wave-powered compute and offers a useful reality-check framework for evaluating ambitious energy-meets-compute claims.
• Does Microsoft’s Clean Energy Pullback Actually Matter? — Examines Microsoft’s shift away from hourly renewable energy matching in its procurement strategy. Relevant for engineers and sustainability professionals working on corporate clean energy procurement, PPAs, and the practical challenges of grid decarbonization claims.

Today’s Synthesis

Combining the open‑source residential demand‑forecasting framework from Forecasting Residential Heating and Electricity Demand with the building‑energy‑management tooling being rolled out by AvalonBay via Parity gives a concrete path for engineers to build real‑time load‑shaping controllers. By feeding high‑resolution probabilistic forecasts into a building’s HVAC and plug‑load schedules, a property can pre‑cool or pre‑heat during low‑carbon windows and then dispatch its flexible demand to the grid when renewable output peaks. Adding a V2G‑capable fleet—such as the vehicles Ford is positioning through Ford Energy —turns the building’s parking garage into a distributed battery that can absorb excess renewable generation and provide ancillary services. An engineer can prototype this loop using the open‑source forecast models, Parity’s API for setpoint control, and a V2G charger’s OCPP interface, delivering a measurable reduction in peak demand and carbon intensity for a commercial portfolio.