AI IN GIS

Why It Matters Across
Today's Industries

The Spatial Shift

From static maps to living, agentic intelligence

GIS is no longer a picture of the world. It is a continuously updating model of it — one that can see, reason, and act on spatial data.

c. 1800 · Engraved map

Past

Static Mapping

Paper sheets and CAD drawings. Surveyed once, outdated soon after, read by humans only.

Sentinel-2 · Satellite

Present

Dynamic GIS

Layered, web-based, 3D. LiDAR, BIM and sensor feeds fused into one queryable spatial model.

Drone · Live view

Now Emerging

Agentic Intelligence

AI that classifies, extracts and decides automatically — turning raw scans into answers.

Showcase 01

Davangere 3D Dashboard

A web platform for automated ward analysis and infrastructure rating — the whole city's condition, ward by ward.

Surface

Potholes · Manholes

Utilities

Electric poles

Underground

Sewage lines

Showcase 02

Revit Automation Plugin

One AI plugin for Scan-to-BIM extraction — from point cloud to Revit elements, in two modes.

1

Human-in-the-loop

Assisted

Engineer guides selection Review each element Maximum control

same plugin

2

Fully Automated

Autonomous

One-click extraction No manual tracing Maximum speed

Part 1 · Human-in-the-loop → Part 2 · Fully Automated

Showcase 03

Custom LiDAR Software

Fully automated AI model classification for complex point clouds — no manual labelling.

Powerline Classification

Conductors and towers isolated from the cloud

Building Classification

Roofs and structures segmented automatically

Showcase 04

Vector Automation Plugin

Automated geometry extraction, written directly to GIS vectors.

Pipeline

Raw Data

AI Detect

GIS Vector

Road Centerlines

Clean network geometry

Polyline Generation

GIS-ready vector output

Future Horizon 01

Agentic GIS & VLM Models

Vision-Language Models driving autonomous spatial agents — systems you can simply ask.

01

See

Read imagery, scans and maps like an expert

02

Understand

Turn plain-language questions into spatial queries

03

Act

Run GIS tools end to end and report back

› "Find every manhole within 50 m of a reported pothole in Ward 12."

Future Horizon 02

Edge AI & Predictive Simulations

Intelligence moves to the field, and the digital twin starts predicting what happens next.

Edge Devices

On-site inference on drones, rovers and field units — results before leaving the site

Dynamic Digital Twins

Spatial simulations of floods, traffic and asset wear — before they happen

Q & A

Let's Build
the Future

Questions, ideas, collaborations — engineers and students, the floor is yours.

3D Dashboards Scan-to-BIM LiDAR AI Vector Automation Agentic GIS Edge AI

Media credits

Title globe: NASA Blue Marble Next Generation — Public domain
Map of Europe, M. Carey / S. Hill (c. 1800) — Public domain
New York City, Sentinel-2 — ESA, attribution
Akureyri drone flight — World&Drone, CC BY 3.0
Potholes in Bengaluru — Mallikarjunasj, CC0
Manhole cover, Viborg — Jebulon, CC0
Varanasi wired area — juggadery, CC BY-SA 2.0
Pipe installation — Tomas Castelazo, CC BY-SA 3.0
Power line pylon near Krishnapur — Kh627, CC BY-SA 3.0
Ouster lidar point cloud, San Francisco — Daniel L. Lu, CC BY 4.0
Autobahnkreuz Köln-Nord — El Grafo, CC BY-SA 3.0
Ontario 401/403/410 interchange — Coolcaesar, CC BY 4.0
NVIDIA Jetson Orin AGX — Auledas, CC BY 4.0
Drone at Mianyang Hi-Tech Expo — RG72, CC BY-SA 4.0
Berlin Mitte — ArildV, CC BY-SA 4.0
Earth at night from the ISS — NASA, Public domain
Videos on slides 7–8 — Google Research, Open Geospatial Solutions, NVIDIA (YouTube)

All photos and clips via Wikimedia Commons.