Edge Manufacturing

December 31, 2019

3D printing has gone through fits and starts, evolving use cases from “everyone can be an inventor” to “shipping at the speed of light”. While customizability is increasingly prized, the truth is that most people aren’t looking to invent every product they use - discovering products that others have put time into thinking about is a wonderful part of being in a society. Even the idea of downloading plans at the speed of light and printing your products immediately hasn’t quite played out as hoped, despite the efficiency argument being quite intuitive: why would we want to slowly move heavy goods and have wasteful packaging when we can just send bits over the wire?

Not That Much Stuff 🔗︎

I’d argue that, despite common tropes about Americans, we don’t actually buy that much stuff that can be 3D printed. We’re not yet at Star Trek Replicator levels yet, and neither Food nor Clothing can be effectively 3D printed at the moment. Many other items such as household supplies, furniture, equipment, etc. require manufacturing processes that are beyond our current generation of additive manufacturing tools.

Given the more limited set of products that can both be 3D printed and are useful, it seems fairly apparent that the capital expenditure for a 3-D printer as well as the maintenance is generally not a worthwhile tradeoff for most individuals. Even the efficiency argument doesn’t pan out that favorably, because you still have to ship the raw materials.

Edge Manufacturing 🔗︎

Given this, I think the adoption curve for 3D printing is mostly going to depend on its ability to enable highly customizable shopping experiences on a product-by-product basis. The core implication of this is that I think the majority of 3D printers are going to be centrally located in the fulfillment centers of a company like Amazon, or in kiosks that are maintained near dense population centers.

Customers will shop online, customize their product, have it manufactured on-demand, and then either have it delivered or pick it up from a nearby kiosk. I’ve taken to calling this “Edge Manufacturing”.

The rationale here is that I believe we’re approaching diminishing returns on shipping latencies, and the next big wave of value for the customer will increasingly come from shopping experiences that allow for extreme levels of customization. It will be merely a side-benefit that as the underlying technology improves, this will come at increasingly attractive economics for an increasingly large range of customizable products (no need for inventory and only need to do last-mile logistics of finished goods).

The key idea is that there are two parts to the experience:

  1. The “Front-end”, which is a product listing experience that allows for a curated customization experience that compiles to CAD.
  2. The “Back-end”, which takes a CAD file and turns it into an actual physical object.

Historically, I believe the focus in 3D printing has been as a distributed “Back-end”, enabling people to invent and manufacture in their own homes. As stated previously, I think the core problem with this is that most people, for most of their problems, are looking to discover products that people have put time and thought into, and when your shopping experience is downloading CAD files, the advantages just don’t outweigh the disadvantages for most.

I believe the path towards a mass-market application lies in the “Front-end” that could exist if we have a Back-end that is as flexible as 3D printers. Here’s an example from my recent experience: I was attempting to organize my apartment, and I wanted to find containers that would allow me to subdivide my drawers and shelves. In a world where SKUs need to be pre-constructed, there is a limited set of dimensions on offer, and of course, none of them fit my furniture exactly how I wanted it to. I have shelf-liners and sub-compartments that aren’t flush. But what if I could simply specify on the product listing, when purchasing the product, the exact dimensions I was looking for by tweaking LxWxH sliders?

There are CAD marketplaces that exist, but here you’re shopping at the level of buying a CAD plan and then tweaking it yourself. This is a complement that makes sense when our mindset is “we want people to own their own 3D printers” - of course they’ll tweak their CAD plans to fit their needs. However, this just doesn’t seem like a realistic shopping experience for most people. What we want is a stand-alone compelling shopping experience with curated customization that compiles to CAD - and that for the vast majority of people, the manufacturing would then be done not on an individual basis but in a fulfillment center or kiosk somewhere else.

2019/12/29 Online Clothes Shopping product