This is what I've done on this project.
The idea was to develop an automated design for CNC machined CNC machine actuator, made from wood-composite panels.
These actuators are intended to be a basic 'block' in a modular CNC machine construction kit.
The design adapts automatically to suit specified motors, ball screws and linear guides. Thicknesses of individual panels and parameters driving the amount of internal reinforcing in the torsion box structure are all user adjustable.
For anyone attempting a design automation project of similar complexity using inventor, I'd recommend avoiding iLogic and just using the API. I've taken the design for this project as far as I can without needing to refactor all the automation code and possibly move to an open source CAD suite, which will let me program with C++ or Python instead of VB.NET.
I made an example actuator using the design to test if all the systems worked as intended. I found my method of using finite element analysis to optimise the structural parameters just didn't work as well as I'd hoped. It takes too long and I have doubts as to the meaningfulness of the results.
Other than that, the idea of using design automation to simplify the development of custom parts for CNC machined CNC machines worked well. Future work needs to look design automation for other modular components, like connecting structures, frameworks and tool holders.
However, this project really got me interested in the idea of developing an algorithm for automatically converting 3D models into structurally function interlocking assemblies made from CNC machined panels. Something that lets you place a 3D object down, apply forces to it, specify panel thicknesses and then automatically generate optimised rigid structures.