Automated Line Inspection
The Challenge
In the railway sector, there is an increasing and urgent challenge related to automating the processes for inspecting and analyzing the condition of railway lines and their installed components.
There is a growing need to introduce intelligent technologies and processes into the maintenance workflow, with the goal of accelerating, automating, tracking, and simplifying a process that is currently managed manually by on-site operators.
Two partners specializing in:
- Software for managing aerial imagery and Artificial Intelligence applied to it
- Railway signaling and control elements
are developing AI-RWay, a platform aimed at automated inspection for monitoring and analyzing the condition of railway infrastructure.
The Project: Innovation for Predictive Maintenance
The project aims to develop an End-to-End online platform to monitor the health of railway infrastructure, analyze its condition, and manage timely notifications, enabling intelligent maintenance by automating the verification and monitoring processes of its components.
This platform includes automated or semi-automated inspections of the railway line, detecting obstacles or anomalies such as rocks, trees, or metallic objects, as well as verifying the condition of signaling equipment, both vertical and horizontal. All of this is made possible through the use of artificial intelligence and the deployment of drones to capture footage and actively monitor the railway line.
By the end of the project, the software platform will include a complete system for:
- Inspecting and analyzing the condition of the railway line with obstacle detection
- Inspecting and analyzing installed components
- Detecting invasive vegetation
This innovative solution is not currently available on the market, addressing a highly interesting and innovative niche.
Platform functionalities
Zirak’s Role
As a project partner, ZIRAK plays an active role in:
- Defining the software architecture of all components (backend, frontend, and data acquisition and communication methods)
- Pre-processing and labeling data collected by drones
- Selecting Artificial Intelligence and Computer Vision models
- Developing the entire software platform
- Managing the on-field demonstrator
The software platform will enable:
- Automatic uploading of drone videos
- Analysis of pre-processed content using AI algorithms tailored to each use case
- Providing outputs with the necessary information to identify detected issues, along with the details required for carrying out the necessary maintenance
- Scheduling interventions, indicating metadata related to the identified issue (e.g., severity), and efficiently managing historical records
To validate the platform effectively, the project includes the creation of a specific dataset for various application use cases, supported by demonstrations in real-world scenarios on identified sections of the reference railway network. These scenarios will include different lighting and weather conditions to ensure better coverage, enabling drone georeferencing as well.
Our partners
AI-RWay is funded as part of the SWIch project, "Sostegno alle attività RSI e alla valorizzazione economica dell’innovazione", programma regionale F.E.S.R. 2021/2027
