Navigating cadCAD’s existence
If you’re not aware of what cadCAD is, it is a modelling and simulation tool for ‘digital-twins’ of complex adaptive systems. It stands for complex adaptive dynamics Computer Aided Design and was developed by Block Science.
In 2019, they open-sourced it, citing that it eliminates the cost of entry for modeling and simulating complex cyber-physical systems. It’s quite popular among the token engineering community.
What are digital twins?
Digital twins are replicas of an organic or inorganic cyber-physical system in a digital form that can be modeled and experimented upon via simulations.
When you are developing a Tesla or a Falcon-9, you have digital replicas of the entire Tesla Model S or the Falcon-9. Commonly digital twins are developed by subsystem and then integrated at system-level. For example, a digital twin of the Merlin engine part of the propulsion subsystem. And the digital twin model maybe (not necessarily) be integrated with the entire Falcon-9 digital twin model. Companies like Valispace, Mathworks and others aid engineers to create these digital twins at various levels of decomposition of their system and for various types of engineering methods (e.g. Ansys for computational fluid dynamics, MATLAB for pretty much most models, Solidworks/CATIA for mechanical design models).
MATLAB from Mathworks has been around pretty long and is popular among engineers in large corporations. Its tools can model and simulate at system-level (depending on what point you define system-level) and also compiles to lower levels for easier deployment into micro controllers.
Valispace on the other hand is an aggregator and works to aggregate sub-system level models so every requirement for every sub-system can be captured and traced as per the high level goal of the development process accounting for tools like MATLAB to be included in its tool set.
The experience so far
Its still early days. It works well for the most part when build systems that can be modeled using differential equations. While, it is trying to be an open-source version of MATLAB. There is still a lot of work to do. A feature like a graphic-based modeling language similar to Simulink would reduce the barriers of learning even further for cadCAD. There is Octave (an open-source alternative to MATLAB), but neither cadCAD or Octave are a match for MATLAB at the time of writing this post.
This is actually terrible if you care about engineering with a Quality Assurance (QA) focus in mind. Why?
Why…
MATLAB has high barriers to entry (way too expensive for a license for the early startup, the hobbyist, hacker, tinkerer, student, in spite of the student licenses) which isn’t good if you want the next Steve Wozniak to invent and innovate something. Especially, crypto-economic systems which can be public digital infrastructure of the future. A popular example of a crypto-economic system that is a public digital infrastructure — Bitcoin.
For the first time in history, anyone can create public digital infrastructure from their bedroom (in their pajamas, if they want) rather than hire contractors, sub-contractor, etc. to build it — like a bridge, a dam or even non-public infrastructure (but really could be) like an aircraft.
Two decades ago they said this about people starting internet companies in their bedrooms/garages in their pajamas. Today there is Shopify, Mailchimp among a plethora of companies (that were probably started in bedrooms/garages) that help people start companies in their bedrooms and garages.
Engineering public digital infrastructure
Now, if you’re building public digital infrastructure, you may want to make sure its tested and QA’d thoroughly. This is where tools like cadCAD, MATLAB, Octave come in handy — especially from a modelling and simulation perspective.
For the first time in history, anyone can create public digital infrastructure from their bedroom.
In traditional (software, hardware, civil, etc.) engineering, there are teams designated to Quality Assurance, to Integration, to Assembly Verification and Testing, to Compliance and more. When engineering firms build that Dam or Bridge or Motorway, there is significant modelling and simulation work done, with some (and many more not mentioned ) of the tools mentioned above, before production.
If anyone in their bedroom can code up the public digital infrastructure of the future, you may want to make sure, the barrier to entry to tools that ensure a properly tested and QA’d crypto-economic system is easily possible prior to production otherwise, they’ll test in prod. This is what cadCAD seems to be doing and that’s exactly what is needed for future public digital infrastructure.
