A.I. Vision & Humanoid Software

NoMines
Technology for a safer world

NoMines is a not-for-profit technology initiative developing an AI-powered decision-support system for the identification of mines, ordnance, and explosive hazards in field conditions. Designed for humanitarian demining specialists, EOD units, and emergency response services.

6,279 Casualties in 2024
60+ Countries affected
<3s Identification time
01 / About the project

Intelligent ordnance identification from a single photo.

NoMines is a mobile AI agent for intelligent identification of munitions, mines, and explosive devices. The system recognizes objects in various states — partially concealed, damaged, or fragmented. Even from a single component, NoMines reconstructs and identifies the full device, delivering a precise structured output: object type, country of origin, production period, hazard class, and handling protocol.

The solution is designed for civilian demining operations, law enforcement, emergency services, and specialized units where accuracy and speed of decision-making are critical.

Not-for-profit humanitarian technology initiative
Computer Vision AI

Multimodal vision models trained on large-scale ordnance datasets including factory samples, field photographs, corroded and deformed objects.

Field-Grade Reliability

Operates offline. Results in under 3 seconds. GPS coordinates logged automatically. Built for demining, law enforcement, and emergency response.

Humanoid Integration Roadmap

Long-term objective: integration into robotic and humanoid systems for autonomous demining in environments where human presence carries unacceptable risk.

Four steps.
Zero ambiguity.

NoMines uses multimodal computer vision models for ordnance identification from visual data. The system accepts an image as input and generates a structured identification output.

01
Capture

Photograph the object in the field. Any angle, any lighting, any condition.

02
AI Analysis

Vision model analyzes morphology, markings, dimensions and surface condition simultaneously.

03
Identification

Type, country, year, hazard class, confidence score, and handling protocol — delivered in under 3 seconds.

04
GPS Log

Coordinates and photo auto-pushed to secure shared database, building a living contamination map.

What the model recognizes

  • Anti-personnel & anti-tank mines
  • Mortar rounds & artillery shells
  • Cluster munitions & submunitions
  • IEDs and booby-trap components
  • Fuzes, tail fins, corroded casings
  • Partially buried & deformed objects

Output per identification

  • Ordnance type & specific designation
  • Country of manufacture & production period
  • Hazard classification (IMAS standard)
  • Model confidence score & visual rationale
  • Standard handling protocol
  • Timestamped GPS record
03 / Mission
NoMines exists at the intersection of AI capability and humanitarian urgency.

According to the Landmine Monitor 2025 report (ICBL), 6,279 people were killed or injured by mines and unexploded ordnance across 52 countries in 2024 — the highest figure since 2020. Approximately 90% of casualties were civilians, nearly half of them children. On average, 17 people per day.

More than 110 million mines remain in the ground across 60+ countries. The Anti-Personnel Mine Ban Convention (Ottawa Treaty, 1997) unites 164 states parties. The United Nations regards demining as an integral element of the humanitarian agenda and a key condition for the recovery of affected territories.

NoMines is developed as a technology tool for reducing these risks. The project is not commercial in the traditional sense — its primary purpose is to support humanitarian demining, law enforcement, and emergency services.

Source: ICBL Landmine Monitor 2025 · nomines.org · California, USA
01

Intelligence in the field

Every specialist deserves real-time, expert-level identification — not a paper manual from the 1990s.

02

Collective knowledge

Each identification may expand the dataset following validation — enabling controlled improvement of model accuracy based on field data.

03

Beyond human limits

The long-term roadmap: integration into autonomous robotic systems operating where human presence is not possible.

04 / Contact

Partner with NoMines.

NoMines welcomes inquiries from humanitarian organizations, EOD units, research institutions, emergency response agencies, and potential partners. To request access, discuss collaboration, or learn more about the project, please use the form or contact us directly.

nomines.org
California, USA

We typically respond within 2 business days. All inquiries are treated as confidential.