My product will be me-assembled and fabricated. It will not be a kit. There will be three PCBs, fabricated off-site, that will include all of the essential components, though a fair amount of handwork will still be required.
What is it that makes one object more appealing than another? What is it that gives it a sense of quality, of finish? What makes something seem durable? It’s very durability? What kinds of design cues signal that? There is some genius in design — in discovering those processes that invest an artifact with meaning — more often than not, they are associated with handwork. The crude of over-construction that accompanies a ‘hand’ engineered artifact communicates the natural durability and care invested in a ‘self’ built artifact.
To this end, the final product will include a CNC-milled, hand-sanded, finished — and weather-proofed — wooden piece. I hope it will avoid the pit-falls of the more or less “disposable” accoutrements that we typically attach to bicycles.
This is not a mass-market product. The cost of the components, alone, places it out of the general marketplace for a bike light (see below). Fortunately (or not), but bicycle products is a place where high-end boutique products thrive. It is not unusual for high-end light systems for mountain bikers or night-commuters to run into the hundreds of dollars. Considering the features and level of finish available in this bicycle light system, that, unfortunately, is the range of pricing that we’re looking at.
Suggested Retail Price
I’m thinking of a retail somewhere north of $150, but I wouldn’t be surprised if it heads over $200. In a certain sense, these will be “limited-edition” lights.
Bill of Materials
Total capital required
This really depends on the quantity of the initial buy. Because — as it currently stands — so much ‘handwork’ will go into each item (I’ll do the assembly myself) — there is little incentive to go big initially. While costs would be brought down significantly by a large electronics purchase, the marginal cost will remain high. While an injection molded or otherwise outsourced design could significantly reduce cost — the aesthetics that I’m looking for are not really compatible with those processes. I think 100 total units is suitably (perhaps impossible) ambitious — but it’s a good, “round” number, as they say. I think they could be sold.
I did not adjust any of my numbers for price breaks (there are some that are significant) — but then again, I did not account for shipping or tax in my bill of materials. At the present calculations, discounting the labor and costs involved in machining all of the parts (free!) and my own work, of course, I need about 10k for 100 units.
Time to market
The items I looked at are in-stock on DigiKey and SparkFun. Most of the raw materials will only be limited by shipping. It is possible that there will be bottlenecks on certain items, like the receiver and tramsitter, but I’d have to call to be certain.
Per-unit assembly time
I don’t want to say anything insane like an hour, but I bet it’ll take that long — at least for the first 20.
Total labor required
So a hundren hours. Or a lot. Too much? Not really. While the CNC steps can be automated, to some extend — my current fabrication plan involves using the waterjet, the bridgeport, and the shopbot. That’s pretty nutty. While I want to maintain that ‘craft’ look, I may need to explore other options ultimately.
Because I really still don’t have a proof-of-concept prototype, especially with respect to mounting, weatherproofing, etc., I really need to do more material research.
I don’t really feel that this is a satisfactory or efficient design. I’ve omitted a number of important elements (fasteners), whose inclusion will really change the dimensions of the assembly. Some smart design decisions, in terms of snap-joints and the like, would make a huge different in assembly and weather-proofing. For example, in my prototype, the back and front are hand-fastened onto the main body. They should snap.
Similarly, though it is aesthetically appealing, there is no reason for there to be two distinct turn signals. This would halve the number of batteries, receivers, and everything else. It would limit the attachment possibilities somewhat, but that is surmountable (and something that I haven’t yet considered closely). Instead, both sides could be powered by one battery, one four-output LED driver, one MCU, and so forth!