Visual Engineering

IdeasHumane Software

When building some physical product the feedback loop between the creation of a new idea and actually testing it is far to high. 3D printers and other technologies have tightened this loop more, but we could do more to make it easier to test hypothesis, while also making engineering more accessible to laypersons without formal rigourous training.

Environment for Experimentation

These ideas build off Bret Victor's Seeing Spaces. The core idea being allowing a rapid amount of options to be tested in a small amount of time. While Victor focuses on the physical spaces, I've chosen to emphasize the digitial. Visual Engineering takes these ideas and allows for individuals to rapidly simulate the attributes of a physical object of their own creation.

Seeing Spaces from Bret Victor on Vimeo.

Allowing people to experiment with their own designs fosters a sense of ownership. If the simulations can be accurate enough and fast enough, a leaderboard for say fastest jet plane, or lowest drag coefficient for a car would be possible. The most important goal of such a platform would be to allow people to get a sense of the tradeoffs made when engineering a product, and being able to make those tradeoffs themselves.

I suspect that very little of true engineering merit would come from such a site, but that is not the goal. The goal is to get people curious about the objects around them and how they were made. Simulation is an incredibly useful tool for people to understand a complex system. This is also why I've chosen to focus on the digital, I'd like people to be able to ask questions of the design. "Would it be better to sweep the wings back? Or is a forward sweep better?" Asking such questions of the physical world is incredibly time and resource intensive, but computational power has gotten to the point that such questions could be asked.

Details vs Generalities

One of the primary questions to consider for such a system is how detailed should it be? Too detailed and it is no longer fun, not to mention taking longer to compute. Too simple and very little can be gleamed from such simulations. This balance is not simple to find.