Abstract Manager

Just over 30 years ago, our industry went through its first major technological transformation with the shift to Computer Aided Drafting (CAD). The industry quickly leveraged CAD and transitioned away from 2D drafting and into 3D modelling. Recently, our industry has recently transitioned into Built Information Modeling (BIM). As we have all started to experience the benefits of a BIM workflow, we now look to the future, and wonder what’s next.

The next major disruption in our workflows will include Artificial Intelligence (AI). We now find ourselves on the verge of using AI in our day-to-day engineering activities. But what is Artificial Intelligence, and how can we leverage its incredible power now? What’s the first step in adoption?

This presentation will answer those questions and dissect the transitional steps leading us from BIM modelling and into AI workflows. We will use a standard roadway element as an example, an OPSS concrete box culvert, to illustrate the design process at each of these four steps during the presentation. Parametric Design, Automation, Generative Design, and finally, Machine Learning and AI. We will take you through the design of a box culvert using these four workflows and demonstrate how implementing these steps into your daily engineering routine will improve your efficiency and drive far greater consistency in our work.

Parametric design will concentrate on how cloud-based workflows will take repetitive operations and allow us to simplify designs by creating common parametric objects, like box culverts, to empower our engineers to drive the design parametrically.

Automation will leverage the power of parametric design, and add powerful scripts and custom programming, to automate the design process even further.

Generative design will then take the automated culvert design, and with the use of algorithms, optimize the culvert's function by analyzing different box culvert variations to generate numerous alternatives. Providing data to a more comprehensive review and endorsement of the preferred optimized alternative.

Lastly, with machine learning, we will discuss how to teach the computer to visually recognize sound engineering culvert designs, and improper designs, to fully automate the design of a box culvert. Thus, providing a glimpse into the future of design using AI.

The presentation will conclude with a look at some of the developments in machine learning and AI on Canadian Transportation Projects.