Following on from my previous post regarding consumers designing their own products, one of the first problems is obviously how consumers will do this. It’s likely that the pioneers will be those who aren’t intimidated by the effort it requires to obtain and then master a 3D CAD package. But given the steep learning curve and the frustrations even experienced designers sometimes encounter when working in CAD, it’s unlikely that any more than a small minority of non-professional designers will be willing to invest the time it takes to become a proficient 3D surfacer.

So starting from the standpoint that helping consumers to design their own products is a worthwhile aim, it becomes important to look at the kind of tools which will enable this. Part of my research will look at developments in CAD software - Google SketchUp, Cosmic Blobs, Teddy etc - and ask whether it will ever be possible to produce stylish products using such ’simple’ tools. But I also want to look at other ways that consumers might be empowered to design products, ways which don’t necessarily imitate the processes that professional industrial designers currently use.

I’ve known about GenoForm, a software package from Genometri, for some time now, since a discussion on Core 77. For the most part it has been presented as a tool for industrial designers, one which allows them to explore design variations, though I think it’s fair to say the reaction on Core 77 was mixed. The main argument against seems to be that generating and filtering design variations is an activity that designers do intuitively, and that GenoForm is a ‘brute force’ method, rather than a creative method, of arriving at the best solution. Nonetheless, it seems to me that an iterative design tool may be valuable to those consumers who lack the skills to design their own product from scratch.

GenoForm works as a plug-in for Solidworks, it also requires Microsoft Excel to be installed in order to generate Solidworks design tables. It is available as a trial license through the link above.In this first round of testing, I’ve used a relatively simple vase form, made by revolving a sketch by 360°. The sketch is a spline, constrained at each end and at its inflection points:

After revolving the spline, a base is added and the surface is knitted, thickened and filleted to create a solid. (Note: this isn’t a requirement of GenoForm, which will work with both solids and un-knitted surfaces):

Opening up GenoForm loads the active model - the one which is currently open in Solidworks. The relatively simple UI allows the ‘Creativity’ to be set - basically the extent to which the software will vary the model’s underlying dimensions - and after pressing ‘Go’ GenoForm will generate variations of the original form:

At this point, GenoForm has varied every dimension in the model’s history tree, to an extent determined by the ‘Creativity’ slider. That includes both sketch dimensions and feature dimensions (such as fillet radii, thickness etc). However GenoForm allows much more detailed control over which dimensions are varied, by giving access to what it describes as the models genes:

Here it’s possible to choose which dimensions are varied and which are fixed at the original value: in the screenshot above I have kept the angle at which the spline meets the base constant, as well as the amount the sketch is revolved (360°, to ensure a watertight vase). The slider on the right also allows fine tuning of the variations as a plus or minus percentage of the original value, such that one dimension could be varied a lot whilst another might be varied only a little, and only by increasing a value, not decreasing it. It quickly becomes apparent also that naming critical dimensions in Solidworks is a good strategy.Having set the genes which I wanted to vary another set of variants were generated. GenoForm allows the ones which look interesting to be saved to an album where they can be filtered or exported back into Solidworks. Unfortunately there’s a problem with the implementation of the album, which is that the image is scaled with the window, meaning it’s difficult to know exactly what shape the object is. It’s not so much a problem in identifying extreme differences between variants, but if you were using the tool to assess small changes I can imagine it would become annoying:

After deciding which variants are interesting, the album can be exported back into Solidworks. GenoForm exports the data as an Excel file, which Solidworks will import as a Design Table:

The variants can then be accessed using Solidworks’ Configuration Manager:

and modified in the same ways as a conventional model, for example by going back into a sketch and tweaking the dimensions which GenoForm has assigned:

Although this was a relatively simple test, a number of things struck me about the way GenoForm works. To begin with I felt that some of the criticisms voiced on Core77 were right, in that this is essentially a tool to replace a designer’s intuition. But as I played around a bit more I was surprised by how GenoForm can actually be used in a creative way, if you choose the parameters to vary carefully. It also quickly became obvious that to play to the strengths of the tool requires an approach during modelling which is usually referred to in best-practice techniques, ie think carefully about which dimensions might be varied later, and the relation of features to one another, rather than just their positioning.In terms of the way s that consumers might use such a tool, I can envisage a design process where the industrial designer decides which parameters of a design are fixed, and which might be changed by a consumer. It’s then quite easy to imagine a web-based tool in which a consumer picks some or all of those parameters and then generates variants, as many times as necessary to reach a product the consumer is happy with. It’s arguable whether this is really design of course, but it does seem a valid way of ensuring uniqueness.To take the investigation to another level, the next stage will be to build a more complex model, and try to really understand the decisions needed to get the required level of variation. I’ll be interested to see whether complexity adds to or subtracts from the creativity of the variants with relation to the original.