In 2024 I joined URBNC3 in Switzerland as a technical co-founder. The product were custom insoles and sandals, 3D-printed to the individuals foot shape. We did this by infering the footshape from pictures taken at various angles, and extrapolating features and custom support adaptations into a print ready mesh of and insole or sandal base.
Normal shoes come with flat insoles - the same for everyone. But no-one has identical feet. Why should we not have personalized support for individual feet? The process was built around capturing multiple photographs of a foot from different angles using a mobile app, reconstructing a 3D model of that foot through an API, then running mesh processing and feature-extraction algorithms to generate a print-ready insole or sandal base. My initial focus was on the part after the model: to convert the foot model into a printable insole or sandal mesh I developed our own internal library as trimesh and others didnot provide the functionality I needed - this was before vibecoding was a thing and I wrote the entire codebase over several nights - . Around this library I built a pipeline that would identify foot poses, features, support needed, and extrapolate that into insole and sandal base meshes that offered custom support, comfort and was ready to be sent to the 3D printer. Further testing print materials, calibrating and automatizing the 3D printing and organizing the 3D printer farm in my office room at home. The entire digital pipeline was eventually bundled into an API, app and a graphical UI for customers and shops to use.
The foundational technology pipeline was completed and entered prototype testing phases. The digital design pipeline sped up the manual design process by a factor of10,000x. While the physical production pipeline was not fully scaled.
During our time we got over 5-figures in pre-sales, raised over 6-figures in funding in Switzerland.
Till this day I reguarly wear the beta-test sandals and I genuinly think they are the most comfortable footwear I own.
