GeoAI to exploit point clouds to inform design, construction & operations | Presentation

GeoAI to exploit point clouds to inform design, construction & operations | Presentation

GeoAI to exploit point clouds to inform design, construction & operations | Presentation

GeoAI to Exploit Point Clouds

Informing Design, Construction

& Operations

Ian Koeppel Global Business Development

Transportation ikoeppel@esri.com

Marcella Faraone Head of BIM&GIS Dpt

m.faraone@fstechnology.it

mailto:ikoeppel@esri.com mailto:ikoeppel@esri.com mailto:m.faraone@fstechnology.it

GIS & BIM Convergence

INFRAWORKS

CIVIL 3DAUTOCAD

ArcGIS

Pro SiteScan

INFOWATER PRO

ArcGIS GeoBIM

BIM COLLABORATE

PRO

INFO360

CONSTRUCTION CLOUD

ArcGIS Online & Enterprise

ArcGIS

Living Atlas Field Maps

Basemaps

DOCS

FORMA TANDEM

ArcGIS

Platform

Apps PlatformCloud

ArcGIS

for AUTOCAD

GIS

BIM

Vision Spatial data is enabling digital transformation

Mission Accelerating Digital Twin creation for the full project lifecycle

with GIS and BIM

© Copyright Esri 2025. All rights reserved. Not to be used, distributed or circulated without authorization. Subject to Autodesk Safe Harbor Statement.

Partnership Update

Exploratory DevelopmentsNew in 2025

• Advancing Connectors + ArcGIS GeoBIM

• Esri Content across Autodesk Portfolio

• Water integrations for Hydraulic Modeling

• ArcGIS Connector for Forma In-Market

Conceptual Brainstorming

• Lifecycle Digital Thread

• AECO Sustainability Applications

• Living Digital Twins for Linear Assets

• Real-world Co-ordinate System Collaboration

• 3D Data Management and Visualization

• AI Models & Workflows to Support AECO

• Connected Systems for Field Data Acquisition

© Copyright Esri 2025. All rights reserved. Not to be used, distributed or circulated without authorization. Subject to Autodesk Safe Harbor Statement.

AI is a Catalyst

Esri & Autodesk alignment continues to advance

. . . Accelerating Insight & Action

AI Agents

Enabling agentic (intelligent)

workflows

AI Framework

Foundational Enterprise

AI Technologies

AI Assistants

Improving Access,

Efficiency & Productivity

GeoAI Design AI

Tools & Models to Assist, Automate, & Augment

Agent to Agent

LLMs

TOOLS & DATA

AGENTS

MODELS

© Copyright Esri 2025. All rights reserved. Not to be used, distributed or circulated without authorization. Subject to Autodesk Safe Harbor Statement.

The Railways infrastructure Digital Models and the GeoAI use case

Roma, 21/10/2025

6

FSTechnology – BIM&GIS Team

Ilaria Mascellani GIS Specialist

Stefano Libianchi BIM Expert

Eleonora Troiani BIM Expert

Marcella Faraone Head BIM&GIS Team

Luca Capuani GIS Specialist

Catello Cascone GIS Specialist

Ferrazza Fabiano GIS Specialist

Mallegni Chiara GIS Specialist

Laura Pesce Head GIS Projects

7

The Railways infrastructure Digital Models Program

8

Analysis of industry best practices, trends, and reference

standards

BIM, GIS, IoT, AI & Analytics, Advanced HMI adoption

Compliance with the national regulations procedure

interne vigenti Rationalization of business

requirements

Enhancement of a heterogeneous application

landscape

Digital Models to support the operation of railway infrastructure

A summary of the program

Trends and industry standards

Technologies Business requirements

Systems in use

Regulatory Framework

The Program aims to develop the essential capabilities for the implementation of digital models of assets, systems, and processes, enabling the adoption of predictive and condition-based maintenance strategies with resulting efficiency benefits. This approach facilitates real-time monitoring mechanisms to support data-driven decision-making and the simulation of complex scenarios, guiding investments toward areas of greatest strategic relevance. All of this is pursued through a shared and unified vision, in which the digital model is recognized as a transversal element across business processes.

Joint Working Group to Implement the Vision

Design of the operating model

Design and Implementation of the Integrated Architectural

Landscape

Design of the Organizational Model

Design of the common data model

Design of the integration with the company systems and the

operation methods

Digital Twin(s) based Vision

9

Analysis of industry best practices, trends, and reference

standards

BIM, GIS, IoT, AI & Analytics, Advanced HMI adoption

Compliance with the national regulations and the internal

procedures

Rationalization of business requirements

Enhancement of a heterogeneous application

landscape

Digital Models to support the operation of railway infrastructure

Key steps

Trends and industry standards

Technologies Business requirements

Systems in use

Regulatory Framework

The Program aims to develop the essential capabilities for the implementation of digital models of assets, systems, and processes, enabling the adoption of predictive and condition-based maintenance strategies with resulting efficiency benefits. This approach facilitates real-time monitoring mechanisms to support data-driven decision-making and the simulation of complex scenarios, guiding investments toward areas of greatest strategic relevance. All of this is pursued through a shared and unified vision, in which the digital model is recognized as a transversal element across business processes.

Joint Working Group to Implement the Vision

Design of the operating model

Design and Implementation of the Integrated Architectural

Landscape

Design of the Organizational Model

Design of the common data model

Design of the integration with the company systems and the

operation methods

Digital Twin(s) based Vision

10

Analysis of industry best practices, trends, and reference

standards

BIM, GIS, IoT, AI & Analytics, Advanced HMI adoption

Compliance with the national regulations and the internal

procedures

Rationalization of business requirements

Enhancement of a heterogeneous application

landscape

Digital Models to support the operation of railway infrastructure

Joint Working Group

The Program aims to develop the essential capabilities for the implementation of digital models of assets, systems, and processes, enabling the adoption of predictive and condition-based maintenance strategies with resulting efficiency benefits. This approach facilitates real-time monitoring mechanisms to support data-driven decision-making and the simulation of complex scenarios, guiding investments toward areas of greatest strategic relevance. All of this is pursued through a shared and unified vision, in which the digital model is recognized as a transversal element across business processes.

Joint Working Group to Implement the Vision

Design of the Organizational Model

Design of the integration with the company systems and the

operation methods

Design and Implementation of the Integrated Architectural

Landscape

Digital Twin(s) based Vision

Trends and industry standards

Technologies Business requirements

Systems in use

Regulatory Framework

Design of the Operating Model Design of the Common Data

Model

11

RAIL TRAFFIC

ASSET MANAGEMENT

EXTERNAL DATA SOURCES

RAIL INFRASTRUCTURE DIGITAL TWIN

BUSINESS OPERATIONS BUSINESS ANALYTICS & INTELLIGENCE

SIMULATIONFORECAST

AI MODELS

DATA ANALYSIS

DECISION MAKINGDASHBOARD

CONTROL ROOM

FIELD SYSTEMS

ENVIRONMENT

METEOROLOGICAL

ACOUSTIC

SEISMIC

HYDROGEOLOGICAL

INVESTMENT (INTEGRATED PLANNING)

PLANNING

PROGRAMMING

REPROGRAMMING

TECH LAB

DATA PRODUCTION ENVIRONMENT

MAINTENANCEENGINEERING

P R

O JE

C T

CD Es

HANDOVER

CONSTRUCTION

DESIGN

TRAFFIC PLANNING

TRAFFIC MANAGEMENT

ASSET MANAGEMENT AND

UTILIZATION

CONTRACT MANAGEMENT

OTHER PROJECTS

LABS

DOCUMENTS & MEDIA ARCHIVE

GEOSPATIAL DATA (GIS)

INTEGRATED INFRASTRUCTURE MODEL & VIEWS

ASSET REGISTRY & LIFECYCLE

ASSET STATUS ALARMS &

NOTIFICATIONS ASSET UTILIZATION

ENVIRONMENTAL MODEL AND DATA

BIM MODELS2D/3D MODELS

D IG

IT A

L T

W IN

A D

M IN

IS TR

A TI

O N

INFRASTRUCTURE NAVIGATION ASSET VIEWER ASSET MONITORING IMMERSIVE HMI

MONITORING (IoT PLATFORM)

SC C,

A C

S, C

TC

Io T

SE N

SO R

S

T E

C H

N O

LO G

IC A

L PL

A N

TS IN

T H

E ST

A TI

O N

S

C O

N ST

U C

TI O

N

SI T

E Io

T

M O

B IL

E D

IA G

N O

ST IC

S

C O

M M

ER CI

A L

TR A

IN

D IA

G N

O ST

IC S

FI X

ED D

IA G

N O

ST IC

S

VISUALISATION AND ELABORATION OF EVENTS AND STATES

EDGE PROCESSING

DIAGNOSTICS

DEVICE AND FW MGMT

INSPECTIONS (IN PERSON, WITH DRONES…)

WORKFORCE

MAINTENANCE PROGRAMMING

MAINTENANCE REGULATIONS AND

STANDARDS

Publication for Railway Undertakings and

Institutional Stakeholders (e.g. EPIR, RINF)

Development

and

Testing

environment

Architectural Landscape

12

MUIF Explorer

Unified Railway Infrastructure Model

13

MUIF Explorer – Unified Railway Infrastructure Model

14

MUIF Explorer – Unified Railway Infrastructure Model

15

Use case

Point clouds and GeoAI

16

Project workflow

MUIF data .laz export .laz to .las conversion

GeoAI Point Cloud

Classification

Classified .las are saved

Classification validated on

New

GeoAI model

Training

the model

Placing the classified .las

Classification revision on

Import and digitisation in

Autodesk Civil3D

Conversion to RECAP format

Upload to ACC Classified Point

Cloud Export

17

Input data

Raw unclassified point clouds

18

Automated classification

• Python script

• Automated classification

19

Manual correction

20

Model training – Toolbox

21

Model training – Toolbox

• Process automation: progressive reduction of manual processing

• Improved precision of the classification of the railway elements

• Significant reduction of the classification time (from 5/8 hours to a few minutes for 300 metres of tracks)

Firts iteration Second iteration Third iteration

22

Central storage in ACC

We used ACC to store all our project files centrally:

• .las scans captured from the plane

23

Central storage in ACC

We used ACC to store all our project files centrally:

• .las scans captured from the plane

• .las scans captured from the train

24

Central storage in ACC

We used ACC to store all our project files centrally:

• .las scans captured from the plane

• .las scans captured from the train

• AutoCAD files

25

Central storage in ACC

We used ACC to store all our project files centrally:

• .las scans captured from the plane

• .las scans captured from the train

• AutoCAD files

• Point clouds published from ReCap

26

Semi-automated BIM model

27

Thank you

Copertine e indici Slide 1 Slide 2 Slide 3 Slide 4

SLIDE fondo scuro Slide 5: The Railways infrastructure Digital Models and the GeoAI use case Slide 6: FSTechnology – BIM&GIS Team Slide 7 Slide 8: A summary of the program Slide 9: Key steps Slide 10: Joint Working Group Slide 11 Slide 12 Slide 13: MUIF Explorer – Unified Railway Infrastructure Model Slide 14 Slide 15 Slide 16: Project workflow Slide 17: Input data Slide 18: Automated classification Slide 19: Manual correction Slide 20: Model training – Toolbox Slide 21: Model training – Toolbox Slide 22: Central storage in ACC Slide 23: Central storage in ACC Slide 24: Central storage in ACC Slide 25: Central storage in ACC Slide 26: Semi-automated BIM model Slide 27


Item Type: pdf