Digital engineering: the Missing Link 1

Digital prototyping allows to make an abstract representation of design that don’t depends on particular materials, manufacturing process, etc. It allows to make a model that unites all main properties independent from the manufacturing process. This was real revolution in design-evolutionary process. To check following things you don’t have to make a real product to check how your product will:

  • look like. You don’t have to cast your chair to check how it will look in walnut, pine, gold, platinum, etc. Photo-realistic renders provide real-looking representation of your model
  • behave. Finite Elements Material Analysis (FEM) helps you to identify if you design is strong enough to carry heavy loads. How this limits will change depend on using of different materials

Higher level of representation allowed to unite all common physical properties and to get rid of dealing with issues of particular embodiments. We use reference books with materials properties to predict the behavior of particular embodiment using deductive reasoning approach.


Let’s illustrate this with beloved sample: a chair.

2016-11-14 10_53_11-Autodesk Fusion 360.png

To make new improved design you have to take existing one and modify it this works good several times, until it takes too much time – like a making of real prototype each time you did this before the digital prototyping.


Parameterization in CAD allows to resolve this issue partially – you are still limited by prior geometry.


Sure you can do this by delegating this routine to AI – it will make for you a bunch of new geometry shapes and check if they fit your requirements. This usually called generative design.

Why can unite all possible designs with a higher level of abstraction, as we did with physical model? This approach known as Systems modelling. This is the highest level of abstraction (if something else won’t be invented later).


This 3-layer approach is well-known and commonly used in software-development and known as a triad “model – algorithm – program”.

  • Model is stands for mathematical model that should be implemented by the algorithm.
  • Algorithm represents the workflow that describes the mathematical abstraction. There are a lot of ways to do this.
  • Program stands for source code on any programming language of one of hundreds (even thousands) languages.

As because we are in digital era now, we cannot hesitate to use triad for design engineering.

Here’s the system model of abstract chair which represents all types of chairs that could be designed.


  • Chair – main system
    • Legs – what the chair stands on
    • Joints1 – what is Seat connected to Legs by
    • Seat – seating plane
    • Joints1 – what is Backrest connected to Seat by
    • Backrest – backrest
  • Rear – is what Sitting Person mostly interacts with chair by.
  • Floor – what the chair stands on. If the chair must survive in aggressive environments, we will add this too, and this will restrict us in material or coverage of chair.

Any system may have quality or quantity restrictions, such as number of legs, or their length. Catalog of system models can make development process commonly reusable and CAD-independent.

Probably you will remember that there is a deep learning technology. Sure, but this is a hybrid of system and physical design approaches. Using it you are training neural network on existing models. Then trained network will provide you design that will fit your requirements. But you are still remaining on the level of physical representation, as because there’s no guaranty that by training you’ll create a complete system model of your design.

See you in next articles.

Leave a comment

Your email address will not be published. Required fields are marked *

One thought on “Digital engineering: the Missing Link