In Praise Of Slow Hardware

In all the discussions I’ve had with hardware makers about their products, one thing is becoming clear: in the end, the cheap part is never cheap. Take a look at this post about a Kickstarter project for example. A maker, Michael Ciuffo, had recently funded a very cool QR code clock that used a simple array of LEDs to display the time in QR code.

He ordered the parts from an online supplier – 500 in total – and begin testing them. In all he saw 38 of the 500 fail in basic tests. In short, his “quick and easy” shipment of components from an inexpensive vendor resulted in a 7.6% failure rate.

“I found out this week that sometimes goods and services purchased in China can be of low quality,” he wrote.

In a similar vein, I once spoke to a hardware broker in Shenzhen who sold bargain-basement phones to the developing world. While his products were far from amazing, he did find similar failure rates in all of the phones he sold, resulting in the need to hire a separate QA tester who powered on and tried all the phones before he shipped them, thereby reducing his profit.

I want to make it clear that this is no jingoistic rant, but this is, in short, the biggest problem with off-shoring hardware manufacturing. However, because the perception is that local – and by local I mean a general U.S. or European audience – is expensive, this quality problem is endlessly repeated.

“When you off-shore hardware, every mistake, and there will be mistakes, causes a delay chain that multiplies by physically shipping prototypes, samples, tester units and more half-way around the world,” said Limor Fried of Adafruit Industries. “One of the best things you can do is keep your supply chain as close as possible.”

It is telling, however, that the company just invested in a $175,000 pick and place machine for their SoHo office.

“This is why we like to manufacture here in SoHo, have our injecting molding in North America, PCBs made in the USA and services like large volume laser cutting here in NYC,” she said.

The proximity of a vendor to your assembly point allows you to, in a pinch, drive to complain. As it stands, Ciuffo’s vendor was kind enough to respond and resend extra pieces but after a 35 day wait on the original LEDs he had already added a month to his build time. While the price of the pieces was obviously low enough for him to consider the opportunity, the cost in time and potentially QA headaches becomes an intangible.

But therein lies the problem: you can’t always source, say, an array of LEDs locally. Chances are the pieces are pulled from the same factory you’d be going to in Shenzhen and, barring a bit of QA on arrival, you might be running into the same problems. However, as companies like Adafruit begin catering to the hobbyist and local manufacturers begin catering to smaller batch hardware creators, I could definitely see it becoming easier to become a true hardware locovore.

We, as consumers, should also require that the things we buy be locally sourced. While I am well aware that manufacturing is not all puppy dogs and rainbows, there is something to be said for a sourcing infrastructure that allows a Kickstarter project lead to make a few calls and flow a bit of money back into the community, state, or country. You either pay for cheap hardware up front or later on, in support costs. An active slow hardware movement would allow far more control over the process of making cool things and would, in the end, benefit us all by raising quality across the board.