Table of Contents
1. Motivation
2. Software diagram / implementation
3. Hardware
4. Integration
5. Tests
6. Problems
7.. Result


Rust is an iron oxide, a usually red oxide formed by the redox reaction of iron and oxygen in the presence of water or air moisture.Rusting is the common term for corrosion of iron and its alloys, such as steel. Many other metals undergo similar corrosion, but the resulting oxides are not commonly called rust.

Cost of corrosion in the infrastructure Industry

A study conducted from 1999 to 2001 shows that a total of $276 billion is spent on corrosion-related issues in commercial, residential and transportation sectors. This cost is approximately 3% of the U.S GDP in the year 1998 (U.S GDP in 1998 was $9.09 trillion). 

Some of the effects of corrosion include a significant deterioration of structural and architectural elements as well as increase the risk of catastrophic equipment failures.






Software diagram Point Cloud generation / Segmentation

Geometrical Segmentation Rust Hotspots

Construction Site Rust Detection

Software diagram Mission Planner

Structural Inspection Path Planning


Mission Planner Preparation
Pointcloud – Mesh – Octomap

Element – Pointcloud – Uniform Mesh


Drone controller

Drone position control

Drone Position Control
    Experiment 1

    Experiment 2

Neural Network – Rust Detection –
    Binary CNN


Region-based Segmentation Using pixel values -threshold Segmentation

Semantic Segmentation

Image Slicing

Neural Network

Localisation – Pointcloud


Construction Site  Rust Detection





Blossomrust is a project of IaaC, Institute for Advanced Architecture of Catalonia developed at Masters of Robotics and Advanced Construction in 2019-2020 by Students: Cedric Droogmans, Luis Jayme Buerba, Abdelrahman Koura & Alexandros Varvantakis, Faculty: Carlos Rizzo & Soroush Garivani.