While the concept of customized 3D-printed bicycle saddles has been around for a few years now, a scalable workflow made for mass customization and serial production has not been possible — until now.

Hyperganic worked with its partners Fizik, the Italian bicycle saddle brand, BASF Forward AM, the German material manufacturer, atum3D, the Dutch 3D Printer manufacturer, and InnovationLab, the German printed and organic electronics expert to make the vision of going from 3D body scans to bike-ready saddles through a fully-automated reality.

We join Peter Davis as he recounts some highlights of the engineering process behind the saddle design workflow.

What is mass customization and why are you doing it for bicycle saddles?

If you are familiar with custom-made dress shirts and shoes, the idea of mass customization takes that process of tailoring close-to-body products to the next level. Using Industrial 3D Printing, it is now technically feasible and financially viable to individualize products and mass produce at scale.

However, to individualize objects for every customer, the design of the resulting products needs to be automated. This is where Hyperganic Core comes into play, a platform for algorithmic engineering.

As for bicycle saddles, their design and material development faced some shortcomings in recent years. Together with the importance of bicycles in future sustainable mobility, it highlighted an opportunity for Hyperganic, together with Fizik and our partners, to dramatically change the way bike saddles are designed and manufactured. We envision a future where cyclists, casual commuters and elite athletes alike, can gain access to saddles that are tailored to their physique and customized to every use.

What are some challenges you face, or foresee one to face, in mass customization?

We start by determining what we want in the final product and work backwards to find out what functions and assets are needed to be captured.

The starting point in the mass customization of bicycle saddles is to digitally create faithful representations of the cyclists’ bodies and translate them into usable information. That is where we used sensors to generate pressure maps that indicate the precise mass and pressure intensities that are applied by the cyclists’ bone structures. This pressure map is our main input.

To seamlessly go from pressure maps to personalized saddle designs, we created a purpose-built application using Hyperganic Core. With its collection of A.I. based engineering tools, it algorithmically designs the thousands of struts that make up the saddles’ lattice structures. In this step, the surface contours and lattice beam thicknesses are automatically modulated according to the input pressure maps on a voxel gaussian scale.

As close-to-body products usually have double curved surfaces, the functional lattice does not “fit” nicely within the shape of the saddle. That is why a layer of uniform aesthetic lattices that conforms to the curved surfaces is added on top of the volume of functional lattice. This combination of different lattice structures is a good example of how Hyperganic Core effortlessly automates processes that were previously manual and time-consuming.

A unique rim is also automatically produced to join along the circumference of the lattice cells once the process above has completed. This rim was designed with a passive mechanism so we could dress the shell of the saddle with the 3D-printed component without glue.

Once we have these steps and the engineered elements in place, we empower the designer or saddle client to pick any lattice structure for the saddle and to apply a pressure map to the lattice structure and the process for one bespoke saddle is completed. In a matter of minutes and seconds, we can head over to customize the next saddle, all without manual interception. Should a tweak in design be needed, we could jump back into Hyperganic Core to tweak certain parameters and run the algorithms again. This allows for rapid iteration of product and accelerates refinement.

The end products are bicycle saddles that minimize peak pressure zones with tailored lattice structures that deliver unprecedented comfort for casual commuters and a performance leap for competitive cyclists.

What are some challenges you face, or foresee one to face, in mass customization?

I think 3D Printing is about the harmony between material, hardware and software. Design, although important, depends on the other factors. This harmony becomes more important when we deal with mass customization because it lays the foundation over which we customize a product.

People would normally start with a desired design and run into problems with material and reiterate the designs only to find out that it is not printable.

In our case, the saddles are printed using Direct Lighting Processing (DLP) and have to be oriented in the right way to avoid warping based on its center of gravity, printing direction and alignment within the build volume of the printer.

That is where the strength of Hyperganic Core lies. You can describe the constraints of a material or a hardware and in turn create a design domain that always results in printable products, whatever the parameters may be.

The other challenge is deciding among the myriad of options we have in going from design file to the printer. With Hyperganic Core, we can export native code directly to the chosen printer.

In our case, after knowing that we will be using atum3D’s DLP technology, we exported the saddles out in bitmap stacks. This unlocks the legacy cap of STL or 3MF file formats and allows us to work in an environment that is in synergy with BASF’s material. In our way, the output is automatically sliced and ready for production.

How do you think Hyperganic unlocks the potential of mass customization?

It really boils down to two key advantages of Hyperganic Core — algorithmic engineering and geometry representation.

Perhaps the most important reason why Hyperganic Core is best suited for mass customization is that it creates objects algorithmically, without the need for manual intervention. Engineers and designers now parametrically frame the challenges with mathematical models and rapidly iterate through various designs as quickly as the computer can create them. This decreases production downtime to a point that makes sense for serial production of mass-customized goods.

The democratization of algorithmic engineering will not just empower engineers, but also end users. Our front-end applications, where the parameters are set, allow users to customize saddles in an environment that is always personal. Turning bespoke products into truly bespoke experiences.

Secondly, how geometries are described through meshes, boundary representations and Non-Uniform Rational B-Splines (NURBs) form bottlenecks when it comes to performing operations such as booleans where underlying mathematical models may struggle to keep up with the ever-increasing complexity.

With Hyperganic Core’s voxel-based geometric kernel, there either is matter at a given point in space, or there isn’t. Operations we perform on geometries cannot fail mathematically and this makes them lightning fast. Volumetric information such as material properties can also be stored if we wish to do multi-color or multi-material printing. This gives rise to unprecedented design freedom that allows us to create products that we could not even comprehend at this moment.

Looking ahead, what else do you think will be mass customized?

I foresee a lot more products to be mass customized in the future. Close-to-body applications like saddles are great low-hanging fruits that can be produced quickly and with high returns on investment. There is also nothing stopping someone from designing a fully automated workflow that customizes entire bicycles based on client data, but what needs to happen first is a paradigm shift — people need to realize that software can drive product development.

This is the paradigm shift that Hyperganic is bringing about, and it will result in us unlocking every printable space for hardware companies, embellishing on the strengths and compliancy of materials produced for this manufacturing revolution.

About Peter

A self-confessed computational nerd, Peter Davis has an extensive background working with artists, companies and research institutes to express their innovative visions. Whether that’s bespoke creations for installations in public spaces to serial production of customized products incorporating client data.

Finishing a career as a professional breakdancer and selling his first company specializing in Urban entertainment, Peter wanted to continue into a new industry with the potential of strong creative expression. This gave rise to co-founding a second company specializing in Computational design, 3D Printing AR/VR and visualizations. After much deliberation their company got acquired by ECCO to start an Innovation lab for the future of leather and shoe production. This bizarre career path has led to Peter pushing the boundaries in all aspects of 3D, robotics, 4.0 manufacturing and even jet packs.

Peter now is spearheading the automation of bespoke products by dramatically accelerating innovation through A.I. based engineering. Products, structures to entire machines are engineered through computer algorithms and mass produced in Digital Factories using Industrial 3D Printing. The resulting objects are more advanced, efficient, and sustainable than conventional products.