Needs
Geosintesi SpA, a leading company in railway vegetation control services, needed to create a technological system capable of recognizing vegetation for the construction of the intelligent train of the RailLandascape project, presented to the European Parliament on 19/03/2019 weed.
Transformation obtained
Particle contributed to the creation of the first intelligent train, the DIC-80, capable of automating the weeding system of railway lines, managing the stabilization of the slopes and the state of the infrastructures.
This highly technological system, designed and built by our professionals, is able to provide all the information relating to the surrounding environment on which it is necessary to intervene, in a selective and differentiated manner through consultation from mobile smartphone or tablet devices. This operating mode allows to intervene on the line in a timely and selective manner.
Work Process
The project was structured on three fundamental work processes, according to a management of developments through AGILE/Scrum with weekly release plans.
First of all, a monitoring system was created that would allow the real-time recognition of the areas that have a range of green color that can be classified as a weed, information necessary to activate the weeding system in an automated and intelligent way. To do this, high-resolution RGB cameras for industrial use were used for the collection of images and their consequent fractioning into analysis areas for the targeted herbicide spraying.
On the vehicle front, we use a Lidar-type laser scanner for the morphological recognition of the area crossed by the railway vehicle. The system analyzes subsequent surveys in compatible areas according to discretionary analyzes that lead to highlighting possible problems such as landslides, misalignment of signs, possible failure of works of art.
Three specific Apps have been created to collect information and surveys.
A Web App for the management of intervention plans and information collection, a Mobile App manageable from smartphones and tablets, for the management of manual interventions and on-site surveys, making the collection of information simple and immediate.
An OnBoard App has been created to detect areas with weeds, which manages all the information acquired while traveling the line.
