3D Geodata Academy

Limited time
3D Forge: Build Camp
Join Newsletter
Login
3D Geodata Academy logo, back to homepage
Enterprise
Start Free
Free · no card · yours to keep

Five free missions on LiDAR point clouds, Python and 3D AI

The Architect's Path is the free opening act of the 3D AI Accelerator. Five videos, 44 minutes in total, each one with a kit you download: the guides, the code pack, and the Neurones 3D software. You answer three questions first, so the plan you print at the end is built around the work you actually want to do.

5 missions, 44 minutesVideos of 9, 9, 9, 9 and 8 minutes, in order.
Python, Open3D, point cloudsVersion pinning, a loader contract, a viewer, a tiny CLI.
A kit with every missionBooklets, guides, a tools pack and a code pack you download.
A rank and a printed planFinish the five and you print your 90 day roadmap.
From 3D data to spatial AI: raw scans become semantic 3D, then scene graphs, then questions a spatial AI agent can answer
The ground the five missions cover: a raw scan carries geometry and colour, semantic 3D gives every point a meaning, a scene graph turns those meanings into relations, and an agent answers questions against it.
Get the five missions → One email address. The link to the mission hub lands in your inbox.

Methods used by teams at AirbusMetaCNRSBMW

Preview: a look inside the free mission
Free mission

Get your free 3D AI mission

A hands-on mission with the data, the code and my walkthrough, sent straight to your inbox.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Act 1 of the 3D AI Accelerator

The five missions, in the order you watch them

Every mission is one video, one kit you download, and one exercise you do on the page itself. You finish a mission to open the next one, and your rank changes at each step.

  1. A scanned building interior classified floor by floor, every room separated by colour
    Rank earned: Cadet · 9 min

    Welcome to the Architect Program

    The six act spine of the whole program, and the six things a 3D AI architect is actually paid to do: measure, align, label, segment, decide, report. The exercise makes you turn over all six before the mission closes.

    KitWelcome Booklet3D AI Architect ebook
  2. An aerial LiDAR point cloud of a city block classified into ground, buildings and vegetation
    Rank earned: Explorer · 9 min

    The 3D Spatial Intelligence Opportunity

    Where the skill gap sits right now, industry by industry, and why hiring managers in mapping, surveying and AEC ask for a decision rather than a model. The exercise builds you a one line offer out of an industry, a spatial outcome and the decision you deliver.

    KitSpatial AI guideInvestment and ROI guide
  3. A laser scan of a working lab, desks and machines captured as a coloured point cloud
    Rank earned: Navigator · 9 min

    The Architect's Setup: a pinned Python environment

    Your environment treated as a specification instead of a laptop. Version pinning and why it matters, with the one I learned the hard way: an Open3D minor bump that silently broke KDTree results on a client job. The exercise makes you pin five packages correctly.

    KitSetup explainerTools packTechnical setup guide
  4. A point cloud of a playground where every structure carries its own segmentation colour
    Rank earned: Operator · 9 min

    The AI Architect Workflow: intent, plan, execute, verify

    The four step loop that keeps you directing a model instead of competing with it. Write the intent sentence before you touch the keyboard, because "segment walls accurate to 2 cm in rooms over 10 square metres" is a different program from "segment walls". The exercise deploys an agent over a warehouse scan and hands back an artifact.

    KitSpatial AI guideBrowser PLY viewer
  5. A scanned library room reconstructed in full colour, shelves and furniture intact
    Rank earned: Architect Padawan · 8 min

    Act 1: building 3D Python systems that survive the week

    Pin, Loader, Viewer, Script, Ship. The five pieces that turn a throwaway notebook into software you can hand over, with the loader as the contract between raw files and everything downstream. Own the loader and you stop chasing the "why is my point cloud grey" ticket. The exercise puts the five stages in order until a file runs the whole pipeline.

    KitLiDAR and Python guideCode packNeurones 3D software

Finish the five and you print two things: a certificate and a 90 day roadmap built from the three questions you answered before Mission 1.

Send me the missions →
What you download with every mission

What you download, and the Python stack it runs on

Every mission carries its own kit, and the kit stays yours whether or not you ever buy anything here.

Read

Booklets and guides

  • The Welcome Booklet, where you pick the deliverable you want to ship
  • The 3D AI Architect ebook
  • The Spatial AI guide, and the investment and ROI guide
  • The LiDAR with Python guide
Install

Setup and tools

  • The setup explainer, on why a pinned environment is a specification
  • The technical setup guide, step by step
  • The tools pack
  • A PLY viewer that runs in the browser
Run

Code and software

  • The code pack from Mission 5: the loader, the viewer and the small command line tool
  • The Neurones 3D software
  • Your printable 90 day roadmap, written from the three questions you answered at the start

Want to read the long form tutorials before you hand over an email address? They are open, with the Python in the page.

Read the free tutorials
After the free missions, the next course

The next step: one full course, or the whole catalogue

The free act ends with a working Python setup and a first pipeline. There are two sensible things to do next.

A cluster of scanned indoor objects floating on black, each one given its own colour and wrapped in its own white box by the detector The next course

3D Object Detection Engine

€197 one time, lifetime access

Where the free missions stop, this one keeps going: from raw LiDAR to clean 3D bounding boxes, with RANSAC, DBSCAN, PCA, KD-Tree and SAM for 3D.

  • 44 lessons, real scans, a detection engine you own
  • The full repository, commercial licence included
  • Fourteen day money-back guarantee
Build the engine → Or see everything first

The complete catalogue

Free 10 pages, PDF

Every course, every program and every subscription on one ladder, with the exact price of each. Nothing is gated and no email is asked for.

  • Free, entry, deep dives, the operating systems
  • The Library and the 3D AI Architect program
Download the catalogue → Prefer it on screen? The same ladder is at learngeodata.eu/pricing.
Stay connected

Reach out or book a call to keep learning

Reach out for tailored support, or book a call to have more information about new courses .

Get in touch
Book a call
Let’s Connect

Send us your questions and ideas and we will define the best path forward

Send us a message
Scroll to Top