Skip to main content
Menu
Book a free intro call FR

Technology & AI for Controlled-Environment Agriculture

Crop Urbanis is developing an expert-reviewed decision-support layer for vertical farms and hydroponic greenhouses. The goal is to transform repeated agronomic, operational and feasibility workflows into structured digital tools that help teams plan, launch and optimise controlled-environment agriculture projects with clearer data and measurable feedback loops.

Discuss a pilot Download technical brief

The Crop Urbanis Intelligence Layer is a controlled-environment agriculture intelligence MVP, developed alongside consulting, research and training rather than replacing them.

It is designed to structure project inputs, retrieve agronomic knowledge, support feasibility modelling, generate SOP and training material, track operational KPIs and assist expert-reviewed diagnostics for greenhouse and vertical-farm teams. AI outputs support professional judgment and are reviewed against crop biology, facility constraints, climate context, market assumptions and operational reality.

Inputs and outputs designed for real CEA projects

The product is structured around the information teams already use during feasibility, design, commissioning, operation and research. The goal is reusable decision support, not generic chat output.

Input

Inputs the system is designed to structure

  1. 01 Site layouts and available growing area
  2. 02 Utility capacity, energy assumptions and water constraints
  3. 03 Crop targets, cultivar choices and production cycles
  4. 04 Historical yield, quality and loss data when available
  5. 05 Nutrient, fertigation, climate and lighting protocols
  6. 06 Labour, CAPEX/OPEX and market assumptions
  7. 07 Pest, disease and post-harvest observations
  8. 08 Supplier, equipment and automation documentation
  9. 09 Regulatory, safety and compliance constraints
  10. 10 Training materials and farm SOPs
Output

Outputs for operators and project teams

  1. 01 Feasibility model inputs and scenario summaries
  2. 02 Crop protocol drafts and SOP updates
  3. 03 KPI dashboards for yield, quality, labour, energy and stability
  4. 04 Expert-reviewed crop and operations diagnostics
  5. 05 Multilingual training material for farm teams
  6. 06 Structured due-diligence packs for investors and developers
  7. 07 Commissioning and scale-up checklists
  8. 08 Data-ready reports for pilots, research partners and suppliers

Pilot-ready modules

Farm and project data model

Structures site layouts, utility capacity, crop targets, budget assumptions, regulations, facility specs and operating constraints into reusable project records.

Feasibility and scenario support

Supports CAPEX/OPEX assumptions, crop planning, production cycles, labour, energy and market scenarios for investor-ready decision packs.

Crop protocol assistant

Helps draft and update crop protocols, fertigation notes, IPM checklists and operational recommendations for expert review.

SOP and training generator

Turns project-specific protocols into multilingual SOPs, checklists, troubleshooting guides and training material for greenhouse and vertical-farm teams.

KPI dashboard layer

Tracks operational indicators such as yield, quality, cycle length, labour, energy, crop losses, issue frequency and optimisation actions.

Computer-vision and sensing validation

Tests whether images, sensors and environmental logs can support crop monitoring, anomaly detection and operational alerts under controlled validation conditions.

Model evaluation and guardrails

Compares AI outputs against agronomic ground truth, expert corrections, field results and acceptance criteria before use in client-facing workflows.

Cloud-ready deployment

Uses cloud infrastructure for secure hosting, structured storage, APIs, dashboards, inference, document processing, monitoring, logging, backups and partner collaboration.

AI and cloud workflows

The MVP is designed around practical workflows rather than AI hype. Each workflow keeps human review, validation and traceability visible.

01

Agronomic knowledge retrieval

Retrieves relevant crop, climate, nutrition, IPM and operations knowledge from curated project materials and validated sources.

Technical detail

Designed for project files, protocols, trial notes and validated agronomic references so teams can compare recommendations against source context.

02

SOP and training generation

Drafts multilingual training guides, checklists and operating procedures for expert review.

Technical detail

Outputs are treated as drafts: wording, crop assumptions and operational steps are checked before teams use them in training or production.

03

Feasibility and scenario support

Structures assumptions for CAPEX/OPEX, crop planning and production scenarios.

Technical detail

Scenario support keeps assumptions explicit so feasibility work can be reviewed by investors, operators and agronomy leads.

04

KPI dashboarding

Turns farm and project information into measurable indicators for post-launch optimisation.

Technical detail

Dashboard logic focuses on operational indicators such as yield, quality, cycle length, labour, energy and recurring issue patterns.

05

Computer vision and sensing validation

Tests image and sensor workflows for crop monitoring, anomaly detection and operational alerts where useful.

Technical detail

Computer-vision and sensing work remains validation-led: use depends on data quality, crop context and clear acceptance criteria.

06

Model evaluation and guardrails

Compares outputs against agronomic ground truth, expert review and field results before client-facing use.

Technical detail

Guardrails are framed around measurable acceptance criteria, expert corrections and traceable feedback from pilot deployments.

Technical architecture under development

The architecture is intentionally modular so pilot workflows can start with project files and expert review, then expand toward dashboards, APIs, inference and sensing integrations as data quality improves.

  1. Data intake Project data + farm files + protocols + sensor/image inputs
  2. Validation Validation and schema mapping
  3. Knowledge Agronomic knowledge base
  4. Inference AI inference + rule-based checks
  5. Review Expert review Case review
  6. Delivery Dashboard / SOP / diagnostics / feasibility report

MVP roadmap and current status

Status is stated conservatively so partners can understand what exists, what is pilot-ready and what needs validation.

Live

Live foundation

Consulting, field agronomy, SOP/training, feasibility, project data collection, research support and operational optimisation.

Pilot

Pilot-ready with selected partners

Data structuring, KPI dashboards, AI-assisted SOP/training, agronomic knowledge retrieval, diagnostics support and model evaluation workflows.

Case review

Validated case by case

Computer vision, sensing integrations, automated alerts and broader deployment depend on data quality, partner scope, crop context and expert-review requirements.

Why cloud and AI infrastructure matters

Cloud and AI infrastructure supports secure hosting, structured data storage, API services, dashboards, document processing, model inference, knowledge retrieval, image and sensor analysis, logging, monitoring, backups, model evaluation and pilot collaboration environments.

  • Human-in-the-loop
  • Pilot-ready workflows
  • EU-aware data handling
  • Expert-reviewed outputs

How the product fits the business

The intelligence layer is being developed as a project-embedded product for selected pilots. It supports consulting, research, training and operator dashboards today, while selected modules may later become managed digital workflows, pilot access packages or licensed decision-support tools for greenhouse operators, vertical-farm developers, suppliers and research partners.

Technical brief for cloud and AI partners

A short downloadable brief summarises the product narrative, architecture, data inputs, outputs, cloud-credit use case, roadmap and truthfulness guardrails for partner or startup-program reviews.

Download English PDF Download French PDF

MVP changelog

A lightweight public changelog records product direction without pretending a finished public SaaS already exists.

June 2026

Website technology layer published

Public narrative, module map, architecture, cloud needs, business model and collaboration path documented.

Private pilot phase

Selected partner workflows

Data structuring, SOP/training generation, dashboards, retrieval and model-evaluation workflows are scoped case by case.

Next validation track

Sensing, image and diagnostic modules

Computer-vision, sensing integrations and automated alerts remain validation tracks tied to data quality and expert-review requirements.

Bring a real CEA workflow to a pilot discussion

For AI, data, cloud or research partnerships, share the crop context, available data, current workflow, infrastructure preferences, validation needs and expected technical outcome.

Contact technology partnerships See field portfolio