Skip to main content
[ AIXA ]R&D

We research beforeanyone asks for it.

Digital twins, Earth observation, computer vision and verifiable traceability. Every line with its declared maturity and what we have not validated yet.

See research areasCollaborate with us
[ 01 ] Areas

Six areas, one foundation

What we build for olive oil is not specific to olive oil. The same models, data and methods solve problems for cities, public services and energy.

Digital twins

Living models of a plot, an industrial plant or a city, connected to measured data.

  • Atlas
  • Agrotwin
  • Coordinated harvest
  • Perfect storm

Earth observation

Sentinel-2, Landsat and LiDAR calibrated against the real ground, not the manual.

  • Harvest forecast
  • Urban heat island
  • Burned area

Vision and spectroscopy

Measuring what the eye cannot see. When no labelled images exist, we generate synthetic scenes with exact labels.

  • Ripeness from the belt
  • NIR
  • Hyperspectral

Agents on verified data

Assistants that query real data, show the query they ran and state the limits of their calculation.

  • Atlas assistant
  • Public services

Trust and traceability

Signatures, verifiable credentials and public anchoring: proofs a third party can check without trusting anyone.

  • Olive oil passport
  • Olearia Sign

Data and visualisation

Scattered official series turned into a coherent base, and our own tools to show them.

  • Albor
Granada in 3D with Sierra Nevada behind, rebuilt from open data
Granada and Sierra Nevada in 3D, rebuilt only from official open data

Featured line · City · TRL 4–5

Atlas, the digital twin of Granada.

An operating system for the city in 3D: every building, street, stop and census section of Granada in one model, with its history and its indicators. You navigate it like a map and query it like an analyst, department by department.

An assistant answers in plain language with real database queries, flies the camera to what it finds, shows the query it ran and says where its calculation ends. Everything comes from official open sources (Cadastre, INE, IECA, Eurostat, MITMA, Copernicus, OpenStreetMap), so the architecture can be replicated in any municipality.

24,156
3D buildings with year, use and height
1.29 M
Statistical values, down to census section
104
Public transport lines with timetables
823 GWh
Estimated annual solar potential, roof by roof
What it does today
Urban planning
3D buildings, city growth since 1900 and a proposal comparator: shadow, Alhambra views and affected neighbours
Mobility
Traffic, the full public transport network, origin-destination trips and walking accessibility
Environment
Green areas and trees, urban heat island, air quality and vegetation change from satellite
Energy
Solar potential of every roof, self-consumption and payback
Housing and economy
Price per m², income and population by census section, jobs and tourism
Risks
Earthquakes since 1998 with time replay, flooding and climate to 2050
Origin-destination trips between districts (MITMA mobility data)
Origin-destination trips between districts (MITMA mobility data)
Proposal comparator: shadow, Alhambra views and surroundings
Proposal comparator: shadow, Alhambra views and surroundings
Vegetation change between two dates with Sentinel-2
Vegetation change between two dates with Sentinel-2
[ 02 ] Olearia lab

From the olive tree to the mill, measured at source

Olearia is our platform for the olive oil chain. Its lab studies how to measure each delivery on arrival, forecast the harvest and coordinate the campaign among cooperative members.

Freshly washed olives moving along the reception belt of a mill TRL 3

Ripeness from the belt

Observe the skin; check the flesh.

A camera over the reception belt to describe the ripeness of each delivery, checked against control cuts. Tested on real public images and synthetic scenes.

Read the article (in Spanish)
Field technician in an olive grove at dawn with a drone over the canopy TRL 4

Harvest forecast

The harvest, before the harvest.

Forecasting the production of each area before picking, with satellite and weather data, and validating it honestly against official forecasts.

Read the article (in Spanish)
Reception hopper unloading olives in a mill TRL 2

Delivery chemistry (NIR)

Composition, before milling.

Near-infrared spectroscopy to estimate the water and oil content of olives at reception, days before the lab result.

Read the article (in Spanish)
Crew picking olives on a hillside grove at sunset TRL 3

Coordinated harvest

The whole harvest, with every member in view.

A 300-member cooperative simulated day by day, with rain, shared crews and milling capacity, to measure how much the campaign improves when the order is coordinated and how to share the gain.

Read the article (in Spanish)
Under study · TRL 1

Questions we have not opened yet

Shelf life per batch

On what date will this specific batch stop being extra virgin, given its packaging and storage temperature?

Oil left in the pomace

How much oil is being lost in the pomace right now, without waiting for the lab?

Mills that learn together

Does a model learn better across several mills, without any of them sharing their data?

Self-adjusting malaxation

Can the mill adjust malaxation temperature and time to the olives coming in?

Olive by olive

Can a camera grade every olive on the belt by ripeness and damage, one by one?

Status as of 30 September 2026. Results from simulation and real public data; no mill pilot yet.

See the lab on olearia.io
[ 03 ] More lines

Other questions we're answering

Earlier-stage lines, each with its maturity. Some will become products; others will show us where the limit is.

Simulated farm robot among olive trees rebuilt from LiDAR
Olive grove TRL 1–2

Agrotwin

Each plot rebuilt in 3D from national LiDAR (PNOA), with canopy height and traversability, to prepare autonomous machinery. 16.7 ha rebuilt and simulated.

Burned area of the Turre fire (Almería) detected with Sentinel-2
Public sector TRL 2–3

Public services

Home-care routes on Andalusia's real road network and burned-area detection from satellite, checked against official perimeters.

Batch verification: after a value is altered, exactly that signature fails
Traceability TRL 1–2

Olive oil passport

Credentials signed by whoever did each step. Every batch is traced back to the grove and forward to the customer; if someone alters a value after signing, exactly that signature fails.

Stacked bar chart rendered with Albor
Platform TRL 1–2

Albor

Open visualisation library with a Rust and WebGPU core: from declarative spec to chart, for web, apps and Python.

Simulation of a solar storm over the Iberian Peninsula
Energy TRL 1–2

Perfect storm

Physical simulation of an extreme solar storm over the Iberian power grid, with wind turbines, satellites and drones.

TRL: the European technology readiness scale, from 1 (basic principle observed) to 9 (system proven in real operation).

[ 04 ] Collaborate

Let's research together

We are open to collaboration. We are looking for companies, public bodies, universities and research centres that want to test, improve and bring these ideas to life. We share what we measure, including when it does not work.

Companies and public bodies

Be a pilot

Test a line in your mill, cooperative, company or city council, in real conditions. We support you and share every result.

I want to be a pilot
Universities and centres

Research with us

Joint projects, theses and method validation. We bring data, infrastructure and code; we publish what we learn together.

Propose a collaboration
Funders and consortia

Let's build a consortium

Regional, national and European R&D&I programmes. We arrive with technology already built and its maturity declared, ready to move from demonstration to market.

Discuss a funding call
Talent

Research from the inside

Internships, bachelor's and master's theses and PhD candidates who want to work with real data on problems that matter.

Write to us

Do you have a problem worth researching that nobody is solving?

Bring your challenge