Prototyping data tools with AI, a case study꞉ Solar + Battery Atlas | Ember

Latest

Data

About

EN

Home

Latest Updates

Prototyping data tools with AI, a case study: Solar + Battery Atlas

Prototyping data tools with AI, a case study: Solar + Battery Atlas

19 Jun 2026

Ember has been experimenting with using artificial intelligence (AI) to expand the scale, depth and usability of our analytical work. AI allows us to move faster from a research question to a working analytical tool, but more importantly it helps us test more scenarios, combine more datasets and generate more granular insights before committing to a full build. Here, we’d like to share the behind the scenes of this work, with one recent example: the Solar + Battery Atlas.

Case study: Solar + Battery Atlas

The Ember Futures team has been using AI to develop a prototype data tool that shows how round-the-clock solar is a reality for much of the world, building on our analysis from last year.

Last year, Ember modelled 12 locations to show that solar and batteries can already provide reliable, high-uptime power in very different parts of the world. For this prototype, we extended that modelling to 5,000 locations globally, giving us much more granular coverage of where the opportunity is strongest.

We then compared the results against maps of population, electricity access, grid reliability and planned fossil capacity. The aim was to move from “is 24/7 solar technically and economically possible?” to “where does it matter most, who could benefit, and which fossil investments are most exposed?”

Solar and batteries are part of the same technology revolution we’re seeing in artificial intelligence. As the Northern Hemisphere marks the summer solstice this week, it seems like a fitting moment to share this beta tool that showcases how 24/7 solar is already cost-effective in much of the world.

Daan Walter<br>‍ Principal

This allowed us to identify more useful “so-whats” for policymakers and investors, including:

Solar is abundant almost everywhere. Solar on suitable land could generate around 125 times today’s global electricity use, and more than 90% of people live where local solar potential is at least 10 times current demand.

Storage turns solar into reliable, high-uptime power. Nine out of ten people live in places where solar-plus-battery systems can reliably supply more than 80% of annual electricity demand. In the sunniest regions, they can reach 99% uptime.

Cost is already competitive for most people. Four in five people can get 80%-uptime solar-plus-battery power for under $100/MWh; for half of humanity, it is under $80/MWh.

The opportunity is greatest where power systems are weakest. Around 760 million people still lack electricity, and close to 2 billion have unreliable grids. Most of this unmet demand is in sunny regions where solar-plus-battery can already beat planned fossil generation on cost.

Planned fossil capacity is increasingly exposed. Of roughly 850 GW of planned coal and gas capacity, about 590 GW is in regions where solar-plus-battery can already deliver 80%-uptime power for under $100/MWh.

The economics keep improving. By 2030, falling solar and battery costs in line with IEA and BNEF outlooks could put 80%-uptime solar-plus-storage below $80/MWh for over 75% of people, and below $100/MWh for nine in ten.

Its goal is to enable policymakers and investors to answer many important questions, including where solar and batteries can already outcompete new fossil infrastructure, where the opportunity is greatest for energy access, and where planned coal and gas capacity is most exposed.

>>> View the beta tool and share your feedback with Daan Walter

With user feedback, we hope to refine this tool further, making it intuitive and useful, before securing funding to move to a full build as an Ember tool.

The team used AI to accelerate simple but time-consuming data workflows and used AI to prototype the user interface, while keeping Ember’s rigorous, human-validated data models at the core.

We’re exploring how AI can speed up the process for creating useful, high-quality data tools to bring the latest data to the hands of decision-makers shaping the electrotech revolution.

Daan Walter<br>‍ Principal

What AI unlocks: de-risking innovation

AI enables us to move rapidly from concept to web tool, without the need for extensive coding. We define the data sources and methodologies, and instruct AI to visualise this in an interactive data tool as per our criteria. Following data validation, we can then publish the prototype to invite user feedback and iterate on the design in a more practical and hands-on way – or fail fast and move to the next concept.

This co-creation process can rapidly lead us to a working prototype, and we can then secure funding and partnerships to turn the ideas that are most useful for policymakers and investors into fully-fledged Ember data tools.

Turning prototypes into Ember-approved data...