AI-built landing and prospecting maps from public Moon mission data
TL;DR Lunarmap turns the Moon data that ISRO and NASA already publish into ready-to-use maps of where to land, where to drive and where ice is likely. Lander and rover companies, research labs and agencies pay per study or per subscription. It needs no spacecraft of its own.
India has excellent lunar data, and hardly anyone can use it quickly. Chandrayaan-2's orbiter carries eight instruments, including a high-resolution camera, an infrared spectrometer and a dual-frequency radar, and its archive is public on ISRO's PRADAN portal in the planetary PDS4 format[1]. In November 2025 ISRO added polar radar mosaics built from about 1,400 datasets at 25 m per pixel, with layers hinting at possible water ice[2].
So what's the catch? Getting to it is slow. The first release came in December 2020 after two missed dates, and portal registration could take weeks[3]. Then merging those files with NASA data and turning them into a slope-safe landing ellipse eats up weeks of specialist work for every single team.
The seed pitched an AI satellite constellation to map resources on other worlds. A private lunar constellation from India would cost far more than any seed round, and the instruments it would carry already fly on government orbiters. So I kept the goal of AI-driven resource maps and dropped the hardware.
Lunarmap ingests public lunar archives, aligns them to one grid and runs models trained on known terrain to produce three products. A landing-site report scores candidate zones on slope, boulders, lighting and radio line of sight. A traverse map plans rover routes. A prospecting layer ranks permanently shadowed craters by how strongly radar, temperature and spectral signals agree on ice, with clear uncertainty.
Every map states its inputs and confidence. Lunarmap does not claim to find ice - it tells you where to look first and why.
Users are mission engineers and planetary scientists. Payers are lander and rover companies such as a US rover maker building a crewed rover for NASA's south-pole missions[4], plus agencies, defence research labs and universities writing proposals. Expect per-study contracts first, sized by area and resolution, with a subscription later for teams that keep coming back to the same sites. Nobody publishes prices for comparable studies.
Chandrayaan-4, the sample-return mission, was approved in September 2024 at ₹2,104.06 Cr for 2028[5], and it needs a safe, scientifically rich landing site. The 2023 space policy lets Indian private firms recover and sell space resources[6]. Commercial lunar firms abroad, from lander makers to resource startups, are planning surface work that starts with site choice.
There is no public figure for the lunar data analytics market. The narrow signal is the data supply - ISRO keeps releasing higher-level lunar products for free[2], which takes your input cost to zero.
The adjacent one is a US lunar-resource startup that says it holds more than US$500M in binding purchase orders for helium-3[7]. That's a buyer that will need prospecting maps first.
ISRO's own MapBrowse tool lets users view Chandrayaan-2 layers but does not score sites[2]. NASA and university teams publish site studies as papers, which are free but slow and not tailored. That startup lists resource harvesting but does not sell maps to others[7]. I found no Indian company selling lunar site analytics, and that's the gap.
| Difficulty | Moderate (40/100) |
| Starting capital | ₹60 L–1.1 Cr |
| Time to launch | 6–9 months |
The software is tractable for a small team with planetary-science depth. The hard part is credibility, because your buyers are a few dozen expert teams worldwide and every one of them will check your method against their own. Ground truth is scarce too, so proving your ice predictions right may take years.
No special licence. Analysing public lunar data is not a space activity that needs IN-SPACe authorisation[6].
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't trust AI alone to pick a landing site. A wrong call loses a mission, so a planetary scientist reviews every ranking and the source data behind it.
Want to work out the AI layer for Lunarmap? Talk to Yash →
I like this for a planetary scientist and an ML engineer who want a lean, low-capital space business and can sell to researchers. If you're expecting consumer-scale revenue then it's the wrong idea for you. Before you build the platform, prove that three mission teams will pay for one landing-site study each.
Plan for roughly ₹60 L–1.1 Cr, mostly for a small science and ML team and computing. There's no hardware to buy, since the data is public.
No IN-SPACe licence is needed for analytics. Follow the PRADAN data terms, credit ISRO and check export rules before selling to foreign defence buyers.
Yes, public lunar data from ISRO and other agencies can be used anywhere. Check each data provider's terms and your own country's export rules.
The free Lunarmap validation kit has the line-by-line capital breakdown, the effort scoring, the data-licence and space-policy checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for mission teams and labs, and 5 assumptions to test. Download the kit →
Thinking about building Lunarmap or something close to it? The make-or-break work is winning the trust of a small, expert buyer group, and I can help you work out how to get in front of them. If you need a hand, you know where to find me. Talk to Yash →
For illustration only. Names are fictional and ideas hypothetical; sourced facts are real but may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.