Slashdot Mirror
Sparse's Story Illustrates the Potholes Faced By Hardware Start-Ups
waderoush (1271548) writes "Hardware is Silicon Valley's new religion. Bits and atoms aren't so different after all, the creed goes; just as the cost and complexity of starting a software company has drastically declined over the last decade, it's now becoming much cheaper and easier to start companies that make physical things. But talk to almost any real hardware company, and you'll discover that the promised land is still some distance away. Sparse, a San Francisco product design startup, learned that the hard way. The company raised $66,000 on Kickstarter for its uber-cool theft-proof bicycle lights, but it took more than a year to deliver the first units to backers, thanks to a string of unforeseen manufacturing and supply-chain snafus. 'We had all the t's crossed and all the i's dotted and still there was a big daily surprise,' says industrial designer Colin Owen, Sparse's co-founder and CEO. Today Sparse is shipping and profitable, with a vision to 'change the face of mobility' for urban cyclists, but its story illustrates just how high the bar still is for aspiring hardware entrepreneurs. Says Owen: 'I wish there was more of a handbook for these things, but the biggest hiccups were very localized and unpredictable.'"
from "no company" to "company delivers a product to customers" is not bad at all.
That's why China already owns the USA's ass in manufacturing. There are too many holes in the manufacturing capability now while in China the place to make that other thing is just down the road - like it used to be in the USA.
For all the marketspeak and fancy looks they're still asking $140 for a 200 lumen light. That's about a half step above terrible. The light I use, which is pretty much the minimum brightness I would consider safe as a "see" and not a "be seen" light, is 900 lumens.
A bullet may have your name on it but splash damage is addressed "To whom it may concern."
I was thinking "looks good", until I saw that this setup uses a dual-headed USB charger that sure looks designed for indoor use only. I'm fine with a fixed battery in my cell phone, tablet, and even laptop, but my bike a) lives outdoors and b) need to accept a spare battery because working lights can be a life-or-death matter.
Nice design, but seriously deficient function.
Daily cockups when manufacturing in China? Why I never!
Simple solution, make your product in the US and create some jobs.
I am also crowdsourcing, an OSHW combined drinking water distiller and thermostat cooker which is energy saving:
Twibright Distillcooker
Its ready to release.
The good thing is that my supply chain was already in place, so all I had to do was increase quantities. I did, however, have to design a simple machine (a jig, basically) to semi-automate a task I had intended to do by hand.
http://igg.me/at/minilaser/ if anyone cares.
Liberty - Security - Laziness - Pick any two.
Hardware is Silicon Valley's new religion. Bits and atoms aren't so different after all, the creed goes; just as the cost and complexity of starting a software company has drastically declined over the last decade, it's now becoming much cheaper and easier to start companies that make physical things. But talk to almost any real hardware company, and you'll discover that the promised land is still some distance away.
No. Hardware is Silicon Valley's founding religion. Software came later and now real hardware startups can not get funding. Sparce's experience shows that even if your development is trivial (no significant R&D) and you don't do any of the manufacturing yourself, it can still be a bumpy road to selling product.
I see no evidence that this is improving. All that has happened is that ambitious hardware startups no longer happen and people are getting excited over hobby scale development that didn't use to make the news. Well, to be fair, Kickstarter has allowed "super hobby" scale developments to take off that used to fall into a no-man's land. They were too small to form a viable business around and yet too big for a couple of guys to pull off in their spare time. Still, this is nowhere near a hardware renaissance. The promise land is not just some distance away. There is little evidence that we are going there.
I've backed a bunch of projects on Kickstarter and Indiegogo. A few completed in the time they expected, most didn't. It didn't bother me that they were late, it bothered me they didn't take this kind of stuff into account when setting expections with the backers.
That's a light for a bicycle. Bikes should not be out after dark anyway. Unless you have an unfulfilled death wish.
NP, ride an e-bike, and have a breakout from your 5.2V rail.
Or if you must ride full-manual, and you habitually carry a spare battery you can habitually carry one with velcro and a micro-usb jack, and strap 'er on when it's time to ride. If you don't habitually carry a spare battery then the idea of swapping a replacement when it's late and dark out is kind of unlikely - where would you get it? Who sells C sells by the late hour?
^see what I did there? ;-)
Maybe they just need to change their business model.
Our hardware (also software heavy) start up was recently acquired for an exponent of our annual gross revenue. The company started ~18 months ago, and we went from idea to Indiegogo to shipping in 12. ALL MADE IN NORTH AMERICA (sure some components were sourced from Asia but design and assembly all in NA). I've done harder things in my life.
Do the people buy apple hardware because there is no cheaper alternative
Correct. There is no cheaper alternative to play books, videos, and apps with Apple DRM, except perhaps an as-is previous-generation device from a pawn shop. And when the iPhone and iPod touch first came out, iTunes Plus hadn't landed yet, and people wanted a phone compatible with their library of FairPlay DRM purchases from what was then called the iTunes Music Store. Finally, the iPhone arrived roughly a year before the HTC Dream, and the iPod touch had a four year lead over Samsung's Galaxy Player, giving people plenty of time to lock themselves into what was then called iPhone OS.
uhm, who writes this tripe?
SILICON VALLEY has always been about hardware.
where do you think the word 'silicon' comes from?
sheesh.
#include <stopped_reading_there.jpg>
--
"It is now safe to switch off your computer."
Silicon Valley is great if you can do something consumer facing and very, very easy. There are great designers in Silicon Valley, but the old-school infrastructure and blue collar understanding is gone. The people who are really good at hardware have moved to different places in the country.
If you want to do anything with any amount of technical difficulty, go to San Diego, Boston, Austin, etc.
I misread that. I was hoping this was a startup that had some innovative, cheap way to repair potholes. Some of us have to deal with some really awful potholes even in June, well past the end of winter.
24 beers in a case, 24 hours in a day. Coincidence? I think not!
$160.00 for an over-designed, under-powered, pretty light made out of a couple of pieces of stamped metal, a lens, and some LED's???
Wow... Hipster bicycle wanks are even bigger suckers than Hipster cult of the Mac wanks...
Oh... Wait. My bad - subset of the same Hipster wank crowd. Carry on.
'We had all the t's crossed and all the i's dotted and still there was a big daily surprise,' says industrial designer Colin Owen...
I'll assume Mt. Owen is just inexperienced, and not outright delusional. There are no inherent problems with hardware manufacturing supply lines that experienced managers can't compensate for. If one vendor flakes, you buy from the second or third source you already lined up. In advance, because you are not stupid/inexperienced/delusional.
The fact that people here are surprised by these "unexpected difficulties" is highly telling.
In the virtual world - which is all the data within any properly functioning computer - one thing is made true by design which isn't true of the real world.
It's very, very simple! *Entropy is conserved*.
IE: Within the computer, things stay exactly where they're left. Nothing degrades, bits stay exactly as they are. (abeit RAID 5 may "lose" one every ~100TB or so).
What this means, is that the second law of thermodynamics *does not apply here*.
When you are used to working in such a domain, the only source of entropy comes in *through you*. (or your users). But either way, the environment doesn't suffer "murphy's law".
When you are building something in the real world, and most especially when you are making something for the first time, or even more meta - making a way of mass making new somethings in order to effect change to the whole world. for the better.... 2nd law of thermo/Murphy's Law applies.
And it applies *hard*. In the real world "No plan survives contact with reality".
IE, you might have thought things through "perfectly" in advance - and this is exactly the skill that works fine for programming, especially programming far divorced from the real world - but it *does not work* in practise.
This is because no theory is exactly true - all have a domain of validity. Which is to say - all are wrong under some particular circumstances.
This means that as good as one might be at planning "reality" in advance, it will never work. We just convince ourselves that we're "sticking to the plan" when we're actually really adapting the plan on the fly. Which is to say, NOT FOLLOWING THE PLAN EXACTLY.
All told, in the real world, doing real value-producing work, such as this start up. IS HARD.
Much harder than any fancy software-only work you might ever do, simply because of thermodynamics, which is to say - The universe does not give a shit. Everything that can happen, probably will at some point. Every mistake that could be made, eventually *will* be made by someone somewhere somehow.
People who think that they're shit hot managers, CEO's and the like, probably have long forgotten what development is REALLY like at the "coal face" where the little details keep causing little unexpected problems daily.
They like to gloss over the problems, after all, those little details *ARE SOMEONE ELSES PROBLEM!*. Else, why have employees at all?
Don't you think it's telling just how few jobs have been taken by robots... which are controlled by computers... which need programming... which must predict the real world in all situations the robot might face, in advance....
Where's the janitor bot? Oh wait, turns out a seemingly complete instruction like "go clean the toilet" is beyond even the most powerful AI's that presently exist.
Hell, just walking around without falling over turns out to be HARD.
So no - hardware is a completely different ball game from software, because it requires actual *real work* which is predominantly comprised of not just the interesting high level stuff, but an absolute shit-load of "re solving" old, boring problems, because something in the real world decayed/changed/broke.
And this is fiddly, hard, boring, stressful, unbelievably time consuming, thankless and possibly the least automate-able work that exists.
Which is why all the smart, yet lazy, people prefer to stay as deep as possible in their virtual worlds. And also why global society can't seem to get it's act together, with all the brightest effectively removing themselves from having actual real effect.
In essence, the difference is that between a comp-sci grad, and a computer engineer who does hardware..
So how to you think the programmer/engineer balance goes in some companies?
Do you really think that the part of MSFT that is responsible for their OS's is comprised of engineers?
How about APPL? Do you think their hardware teams are comprised only of really bright programmers?
The summary made it sound like an electronic hardware startup, and the difficulties behind competing with the bigwigs like IBM, AMD, Intel, Cisco, Apple, etc.
No, it's click-bait.
As nice as a bicycle headlamp is (that will still be stolen -- thieves don't use normal tools, and they're usually supporting a drug habit), the article didn't even talk about manufacturing in Silicon Valley, or even San Francisco (which is 45 miles north), and they had no unoriginal issues with certifications in other countries. I'm voting this article down. Re-submit with an accurate summary next time.
Who sells C sells by the late hour?
^see what I did there? ;-)
Yes. You misspelled "cells".
If you think I voted for Trump because of this post, you're wrong. I voted for Dr. Jill Stein of the Green Party. Again.
The right approach to theft is to have a light that's so ubiquituous and cheap that it's not worth stealing - as in: nobody will buy it. Going high-tech, high-pricetag is precisely the wrong thing to do. You should make it not worth stealing in the first place.
from investing in hardware companies.
It sounds like one big excuse for delivering late. They should have anticipated hiccups. That's the case for software companies and web services, too. You build a website that looks great and works great, now what? how to get critical mass? surprises start showing up...
I read the entire article and there isn't a single mention of a specific problem that they faced. It's all.. oh oh it's so hard because surprises showed up. What surprise???
The closest I can find is this:
"The analogy I would make is to when you first futz around with 3D modeling software. It’s joyful and fun to do something, anything. But the frustration comes when you want to do a particular thing. You want a representation of what is in your mind, with specificity and correctness. And that’s very hard to achieve, for random reasons"
Seriously, the reason is random? You can say about software, too. You build a website, and it comes out too slow for random reasons. You develop some state of the art algorithm, and it doesn't run on the hardware in reasonable time for random reasons. Let's face it, you can't go in and make a few tweaks to make the algorithm run faster.
Let's face it. Business is hard. Hardware is not especially harder than any other business. So called "social networking" type businesses tend to pick up faster for various reasons, but the risk is still high.
This article is one huge whiny marketing gimmick.... and deters investors from investing in hardware start-ups. Why are they trying to do a disservice to other entrepreneurs working on hardware ideas?
Pick a field outside of the military (or even inside the military with rocket engines coming from Russia) and it is a shambles, especially with computer and electronic equipment. Try getting something done with a "US" manufacturer when there is a holiday in China or a big snowstorm over there and you'll see exactly how much of a shambles it is. Not a complete failure but some stupid outsourcing choices have removed a competitive edge and some industries have to rely on government enforced trade barriers to survive at all. Personally I see it as due to seeing the ideal as people like Edsel Ford instead of Henry Ford - the cult of the specialist manager with no clue about what they are managing dropped into a "too big to fail" situation and seen as a success. Today even people like Donald Trump with a vastly negative net worth are seen as winners because they pinned the consequences of their multiple failures on others. You can't run things so badly without consequences, hence the slip to number two and a continued decline.
Too many of these supposed "high tech hardware startups" are producing the kind of crap that came from China two decades ago and Japan four decades ago. Bicycle lights. iPhone cases. Even the Raspberry Pi and the Arduino are just PC boards stuck under systems on a chip made in China. This is not high tech.
There were some guys at TechShop last year making a plastic gizmo for attaching an iWhatever to a an auto dashboard. They had a big "Made in Silicon Valley" poster. I felt they were embarassing Silicon Valley.
We need to do better than this.
the biggest hiccups were very localized and unpredictable.
What a surprise.
The things you anticipate are those that you predicted and prepared for. It is always the unpredicted ones which cause hiccups.
In the end, you cannot prepare for all eventualities, but you must budget for a number of them that will hit you, even when you cannot say precisely in advance what or when they will be. If you don't, your project will come in late and over budget.
You know it's time for the next revolution when your rulers' names end with roman numerals.
TFA is 20 full-screen pictures of their product, and page after page of copy about how awesome the product is. Only barely a mention of some minor hiccups, that get treated as an industry problem, rather than the realities of an incompetent start-up that simply didn't know WTF it was doing.
And frankly, $140 for a set of 'sleek' bicycle lights makes me want to go on a killing spree.
Buy a couple 3-mode SK68 lights for $5/ea. Brighter than you could ever want, with high/low/strobe, and multiple zoom settings:
http://www.amazon.com/dp/B006E...
Some $2 bike mounts:
http://www.amazon.com/dp/B00AD...
And if you don't want to cut-out some red cellophane to fit, you can get a kit with red lens for the tail light:
http://www.amazon.com/dp/B0000...
Batteries and Charger, < $13:
http://www.amazon.com//dp/B004...
http://www.amazon.com/dp/B004N...
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
https://www.youtube.com/watch?... This is what I use for the machining.
https://www.youtube.com/watch?... This is how I get the 3D printer to just keep going (I am going to be selling this as a kit too).
This is an integral part of engineering a product to me - you can't just make a prototype and send the drawings "off in the cloud" to be made. I mean, I guess you can, but then how can you be sure that it was made well? See all the product recalls, etc.
Liberty - Security - Laziness - Pick any two.
Pick a field outside of the military (or even inside the military with rocket engines coming from Russia) and it is a shambles, especially with computer and electronic equipment.
mac pro made in america. moto made in america. model s made in america. prediction, 10 years more than half of apple's stuff will be made in america.
yeah but if your light is detachable, then you can take it inside to charge it or whatever MIND == BLOWN
Staying anonymous... I'm involved in a hardware start-up. IMNSHO, these guys are diddling around and should get better/professional advice. Our design is very involved and the manufacturing is moderately complex (not even considering the software). It sells for thousands of dollars and we're cued up to sell close to a thousand units this year. In the back of our minds is that the likes of Panasonic, Sanyo, and LG are taking a wait-and-see. We're using US-based tech manufacturer(s) because we get their experience, expertise, contacts, while cutting out timezone and language issues. This expertise costs but can't be underestimated, these manufacturers have people who can look at your design and tear apart your prototype and help you scale the assembly. Now back to time tables, we brainstormed what could go wrong and built into the plan weeks of extra slack for us, the manufacturers, our suppliers, and their suppliers: Holidays in China, MOQ's and stocks, upstream supply problems, substitutions, assembly line revisions, QA testing, more QA testing, agency certifications, storage and delivery, etc. and padded it out a little to cover unknowns. We figured we'd be early. You know what? Even with that padding we're running right up to our Production target date. Why? We didn't pad for ENOUGH unknowns of the forehead slapping variety: supplier fails to include part of an order then puts the components on a slow boat instead of expediting it; assembler ignores BOM and uses wrong component requiring rework; part is in practice completely below datasheet spec, the list goes on. These guys didn't do their homework and didn't ask enough adults what to bring to their first rodeo.
The can do the job. One-person operation. No need to leave the cab to patch a pothole. A patching job takes only a few minutes. Good quality patches. Requires a skilled operator who's good with a complex joystick, controlling air jets, asphalt conveyors, and rollers.
Do web design, and want to do something useful? Contact the company, in Saskatchewan, and offer to redesign their 2009 web site with bad layout, non-streaming video, and a lack of testimonials.
A more ambitious plan would be to use computer vision and robotic control on this thing. It needs an aimbot. Something where the operator aims the boom at the pothole, a Kinect-type sensor maps the pothole, and the computers handle the job of cleaning out the edges of the hole, dumping the right amount of asphalt in the right places, and tamping it down. Then you'd get a consistently good job even with a mediocre operator.
The C store sells C sells by the late hour.
Hardware _is_ Silicon Valley's only religion. Software is a by-product. These guys do not know why it is called *Silicon Valley*.
As I wrote, treading water :(
Hopefully that outlier will become a trend and hopefully it will take less than ten years while some electronics industry still exists in the USA to take up the work.
Fifteen years from now those Chinese made parts will be manufactured in Africa and SE Asia.
227-3517
dude, america's back. we're the factory of the world. apple is just the start. they have the money to overcome the first hurdles, then we start to have a growing trained expert workforce. america you get high technology high quality products mostly automated by robots.
I wish you were correct instead of overconfidently deluded.
heavy manufacturing is the cornerstone of the nation. can't spell america without "I", "me", and "a car"
Heavy manufacturing in the USA has been utterly fucked over by a mismanaged steel industry on government life support that has forgotten how to stand on it's own. That has forced serious costs on other industries and started the avalanche of manufacturing industry departing offshore in the first place some decades back. Electronics is a late departure but most of the heavy stuff left the building about when Elvis did. Think about things like shipbuilding, mining equipment, trains etc - all stuff that used to be made in the USA in vast quantities and exported some years ago but now there is very little of that or any heavy engineering in comparison to the 1970s.