Industry Talk
Illuminating Dark Data – using machine learning to search our drawings
クラス ID: BES500067
製品: AutoCAD, Revit
トピック: Artificial Intelligence (AI), Digital Transformation, Future of Work


As the capabilities of Machine Learning advanced, we had a hunch we could use Machine Learning to unlock the latent potential in decades of GHD Projects 2D drawings. Enabling the ability to search unstructured data stored on network drives and shining light on the knowledge of the past, and ensuring it continues to be available into the future. Like many, the challenge GHD faced was firstly, how do we bring that data into the light, and secondly what real value does it offer. Through an Autodesk Residency Program, our technical applications innovation team set out to explore, using artificial intelligence and machine learning, the volumes of dark data hidden within years of client and project archives. The starting point was the idea of “imagine if we could activate our historical archive of drawings”. So, the first step was to determine if it was possible. Join us to look at how we explored the possibility and what we did when we discovered it was.


  1. Explore creating a search function to unlock the valuable IP in historical 2D engineering drawings, using machine learning.
  2. Anticipate the challenges and benefits of using machine learning to search historical engineering drawings.
  3. Identify what technical capabilities and human skill sets are required to search dark data.
  4. Explore skill set required to use Machine Learning