About Farmonaut and our work with farms and mines

The problem we started with, and why we chose satellites

Farmonaut began in 2018 with a practical problem. Soil sensors, drones and regular agronomist visits all cost money per field, which puts them out of reach for a lot of growers, and even a careful farmer can only walk part of a field in a day. Problems such as water stress or a nutrient shortfall often show up in one corner of a field long before anyone walks past it.

Satellites already pass over every farm on a regular schedule. The European Sentinel-2 mission, for example, images the same spot about every five days with its two satellites and records visible and near-infrared light at 10-meter resolution, which is enough to see differences inside a field. Radar satellites add readings on cloudy days, because radar sees through cloud.

Getting the imagery was the easy part. Most of the effort went into turning a vegetation index into something a grower can act on, such as a water amount in millimeters or a fertilizer dose in kilograms, in a language they read every day. That's still most of the work we do. It's why the field report now comes in 248 languages, runs on a phone, and says "apply nothing" when a field doesn't need any fertilizer at all.

How the same technology serves farms and mines

A crop under stress and a patch of altered rock both change the way the ground reflects and emits light. We first wrote our pipeline for optical, radar, thermal and terrain data to monitor crops, and in 2025 we retuned it for geology. What changes between the two is the model on top and the sensors we lean on.

For farms, we watch the same field again and again through the season, and the value comes from comparing dates: whether a block is drying out faster than the one next to it, or whether the canopy has stopped growing. For mining, we analyze a license in depth, combining eight sensor families that include ASTER shortwave infrared for alteration minerals, radar for faults and lineaments, and EMAG2 magnetics. Multi-season composites help where vegetation hides the rock. The result is a ranked list of Exploration Targets on our satellite mineral detection service, which teams can then explore in Mining Pro.

How we work with customers

On a farm

Most people start in the web app by searching an address, drawing a field or uploading a KML, KMZ or shapefile. Readings arrive after the next satellite pass. From then on, each pass updates 16 satellite views of the field and the AI farm advisory turns them into a report on crop health, water, soil, yield, pests and fertilizer. You can download it as a PDF or share it by link with an agronomist or buyer who doesn't have an account.

For agribusinesses, lenders and software teams

Larger customers usually add many fields at once and want the data inside their own systems. Agronomists keep their clients' fields in one account, field teams get tasks through the Agro Admin app, lenders and insurers use crop loan and insurance checks, and developers pull the same data through the agriculture API. Seven partners run Farmonaut under their own brand.

On a mining license

A mining engagement starts with a boundary sent through the mining quote form. We reply within one working day with the scope, price and timeline, and deliver under 5 days from a usable boundary (larger portfolios are delivered in phases). After delivery, clients take the ranked zones to the field and can keep working with the results in Mining Pro.

How we check our results

On the farming side, the results we publish are customers' own estimates, given through in-app feedback. Before one goes on our site, we compare it with the reports the system actually sent that customer. When John Vansickle, a blueberry farmer in Citrus County, Florida, told us he had cut irrigation and fertilizer use by 30–50%, we could see in his 1,048 reports that in one week the advice was "no water" on 78 of 175 block-days and no fertilizer on 8 of his 25 blocks. You can read that and the other write-ups on the farming case studies page.

In mining, a ranked zone is only confirmed when someone samples or drills it. Clients have taken five engagements to the field or drill so far, and on the engagements where clients tested our ranked zones, 90% were confirmed by ground or drill evidence. On one copper-cobalt license, the client drilled our targets and reported mineralization where the analysis pointed. Five engagements is still a small number, which is one more reason we describe the output as Exploration Targets that come ahead of a JORC or NI 43-101 estimate. Our mining case studies show what clients did next.

What our principles look like in practice

Here is how those principles show up on real accounts.

• A 190-acre farm in El Bayadh, Algeria, received "do not water" on 17 of 28 field-days in one week, with 40 acres off all week, and 3 of its 4 fields needed no fertilizer. The grower estimates 51% less water than watering daily.
• AGROEX Consultancy, working across three countries, received 181 reports in 75 days on 15 fields and 39 warnings above 80% risk. Each pest warning came with its probability, so the team could see which ones to scout first.
• On a gold portfolio in a Zimbabwe greenstone belt, we set aside 83.7% of the area before any fieldwork and narrowed the rest down to seven zones.
• We publish mining figures with the qualifier attached. The 90% confirmation rate always comes with "where clients tested them", and demonstration data in Mining Pro is labeled as conceptual and synthetic.

Where the platform is used

Farmonaut has fields in 135 countries. Apple orchards are monitored in 45 countries, and oil palm plantations account for 116,000+ acres across 35,300+ plots in 11 countries. The mining side has analyzed licenses in 33 countries over this and the last financial year, and Mining Pro has users in 25+ countries.

Agriculture, by region	Acres monitored	Mining, largest countries	Acres analyzed
Asia	853,000	Ghana	144,020
Africa & Middle East	56,800	Tanzania	45,697
Latin America	27,600	Senegal	38,170
North America	20,200	Mauritania	34,842
Oceania	15,600	DR Congo	20,292
Europe	14,100	Mongolia	17,569
		Zambia	16,969
		USA	10,255

Agriculture figures are platform totals. Mining figures cover FY2025-26 and FY2026-27 to date.