The Burro light would also differ in other ways. “Patrick’s original Sun King had three brightness settings,” said Whit. “I wanted four. He was resistant to that, and he knows that stuff so well, but I was pretty adamant. So with strong input from me we came up with four power settings, each with some element critical to our market positioning. We wanted one setting that matched the perceived output of a kerosene hurricane lamp, so people could do the same shit they do with kerosene but at a huge cost savings. Lots of people talk about beating kerosene, but we truly deliver it in this mode—eighty percent cost savings, cleaner, safer, more convenient light. It’s huge for anyone with access to the Burro battery service.
“We wanted a second mode that blew kerosene out of the water—you know, any observer would say that’s brighter than kerosene, but still at a thirty percent lower cost than kerosene.
“Then we were both in agreement that we wanted something really bright. Exactly how bright was the question. I said, ‘How high can this thing go?’ Patrick was like, ‘The human eye isn’t linear, I wouldn’t recommend pushing it all the way; buy yourself a little more battery life and go dimmer, people won’t notice it all that much.’ I said, ‘You know what? Go for it. What’s the maximum current?’ He said, ‘Three hundred and fifty milli-amps.’ He was saying we should go for like three hundred or two eighty. So we did the math on it, and for any of these currents, even assuming typical degradation of our batteries, we’d get maybe only five or six hours of life. I was like, ‘Dude, that’s an evening. Let’s go all the way. Give ’em the brightest fucking light in the village! They’ll get it—they’ll know it’s gonna use more power. Plus they’ll still get another five or so hours in the lower modes after the batteries run low on juice.’ I talked him into it.”
“So that’s three settings,” I said. “You’ve got kerosene equivalent for a fraction of kerosene’s operating cost, better than kerosene for still less money than kerosene, and bright as physically possible for five hours. What’s the fourth setting?”
“Well, I told Patrick how we’re facing this competitive threat from Tiger Head on the insanely long tail they deliver in those stupid flashlights. So I asked him how low we could go and still get some useful light—like an ambient night-light or something like that.”
“What did he say?”
“He said, ‘Try it yourself. Buy an LED and just drive it.’ He explained how to dial in the input voltage and use different resistors to play with different ranges of current. I was home in Seattle then, so I went to Radio Shack and bought an LED and some resistors. I’m doing all this and looking at the output, blocking out the windows, turning out the lights, and I email him and say, ‘Okay, I can go as low as six milli-amps and still get useful light.’ He says, ‘Really?’ I say, ‘Yeah, I mean I guess there’s maybe a photon or two there. But I’m looking at these Tiger Head flashlights after a hundred hours of use with Tiger Head batteries, and that’s what they look like! It’s a really low bar. It’s like a candle. You’re not going out hunting with this light, but there is a glow there, and I’ve learned that villagers like a night-light.’
“Patrick said if we went that low we’d have to factor in the circuit, because the circuit itself would draw as much current as the LED at that point. I actually wanted to go lower. I did a lot of tests, and you could argue that a Tiger Head flashlight with a pair of Tiger Head batteries was maybe usable as a night-light for two hundred or two hundred and fifty hours. So I wanted to beat that, to hit three hundred hours. But we couldn’t really go below six milli-amps, which would give us two hundred-plus hours once we factored in the draw of the circuit, so that was it.”
Whit and Patrick settled on an initial order of three thousand lights and worked together to dial down costs as low as possible. As the project advanced and estimates of manufacturing, shipping, tax, and other import costs firmed up, it became clear that Burro could not hit Whit’s ideal retail price of ten cedis (about seven dollars) while maintaining adequate margins. Even using sea freight, which was dramatically cheaper than air, wouldn’t cut it. It would have to sell for fifteen cedis, roughly ten dollars. But Whit felt the market could bear the price, given the huge cost savings over kerosene.
A hand sample of the light was manufactured and sent to Ghana in September. Whit responded with a detailed email to Patrick, who continued the conversation in a series of answers to Whit’s questions:
From: T. Patrick Walsh
Sent: 9/18/2010 10:03:23 AM
To: Whit Alexander
Subject: Re: Fwd: DHL sample
WA: Couple early thoughts/questions, Patrick: We are seeing flicker really only when first entering any given power mode. Is this the extent of the issue, or is it impacting in ways we aren’t yet noticing?
PW: That’s the extent of the issue.
WA: Battery lid does not seem excessively tight in sample. Was it fixed by hand in the sample?
PW: Yes.
WA: Battery lid is able to be closed in four different orientations, three of which prevent switch from functioning. Not a showstopper, and I wouldn’t want to delay initial order to address this, but should we consider “keying” the lid somehow and/or marking correct orientation?
PW: Already being done. If you look on the back side of the cover there is a stopper that is supposed to prevent the cover from being inserted in the wrong direction. This was designed poorly so it doesn’t work, we have already started altering the tooling to fix it.
WA: Is there added schedule risk if we wanted to go slightly cooler on the light color? Not sure we want this yet, and not sure if it’s all or nothing. Some early feedback that the color combined with the frosting is creating an impression that we’re an “old-fashioned” incandescent. It may be the kind of thing where usage preference is different than purchase preference. Is slightly cooler possible, or is a major shift the only option? Would there be any supply or functionality risk from making such a change so late in the game?
PW: Interesting. We could try to go with some middle-ground between what you have and the Sun King—maybe 4500K–5000K. There will always be a range of colors in each batch of LEDs, so these are ballpark figures. Let me know if that’s what you want to do based on feedback there. There is always some supply risk with LED chips, sometimes you just can’t get exactly the bins you want. But even if we can’t get the exact target bin we can always get something close, and the lead times have become very reasonable nowadays.
WA: Will wire stand be identical to Sun King in shape? The new lamp is slightly wider. Old stand seems to function fine with new lamp, but it does change the “bearing” angle a bit. If the stand was changing, I’d consider trying to make the wall-mounting screw/ nail affordance more pronounced.
PW: The regular Sun King and the AA version are actually the same outer dimensions, so we should keep the exact same stand at least for now, as producing this part in low volume is not trivial.
WA: The new lamp is, no surprise, pretty darn heavy when loaded up with three cells. I’d like to reduce the prospects of lamps crashing to the floor if practical. On that topic, how do you think we’ll perform in drop tests when loaded with cells? I managed a small such test by accident already, and nothing bad happened, although the battery lid did pop off.
PW: This is what we have seen: battery cover pops off if the lamp is dropped on the top side. The mechanism is pretty clear—the batteries hit the cover and push it out. I have done a few drop tests with batteries loaded and even spiked it into the ground touchdown-style and ended up with a blown-off cover but no permanent damage outside of a nick or two. We could try to add more of an undercut to the bosses that hold on the battery cover, but that may be a bad idea as if they are too undercut and the batteries really want out, they could break the bosses right off instead of simply pushing the cover off without damaging it and letting the energy dissipate. Let me know what you think.
WA: What’s been your experience with warranty, Patrick? We’d like to of
fer one year, but we’d want some backing from you guys.
PW: In India we offer a 1-year warranty against manufacturing defects. We also offer the same to international customers even for the original Sun King. The question is how to implement this in practice. If you have defects that lead to returns we will replace the units for sure, we just need to make sure they are actually defective and not abused. Standard warranty against defects in material/ workmanship only, obviously.
WA: What caveats do you have to make and how do you know if someone dropped his lamp into a pond? Should we consider any field servicing? Should we order any spare parts? On branding, the seam between the battery lid and the back is so smooth, we actually might be able to pad print over the two when assembled. This would give us some good real estate upper and lower right for logos. It could also provide that visual hint on correct battery lid orientation.
PW: I don’t see why not!
WA: We’re still playing with our baby all weekend and are likely to have further insights or questions, but this should get us going. We want to order on Monday, so let’s resolve anything we need to do to get you guys unblocked and into manufacturing ASAP.
PW: OK, we should be ready to quote and take your order by Sept. 25th. If we can do so by then we will get this on a plane certainly by the end of October.
WA: Thanks again, Patrick.
PW: My pleasure!
I’m old enough to remember the TV repairman (ours was named, perfectly, Ray) making house calls, a concept even more anachronistic than doctors knocking on your door (I remember that too) since at least doctors still exist. In Ghana, however, TV and radio repair shops are everywhere—wooden shacks with mountains of electronic carcasses overflowing onto the street and where men labor six days a week over circuit boards with tweezers and soldering irons.* Nothing in Africa gets thrown away, because there is no money to buy new, so Africans have learned how to repair just about anything. Ghana’s manufacturing sector may be sadly underdeveloped, but its knowledge base on how stuff works, based on the country’s vibrant repair business, is profound.
The console television that Ray the Repairman serviced in our living room was made in America. It seems reasonable to observe a correlation, and perhaps causality, between a society’s ability to fix things and its ability to make things. Could it be mere coincidence that our throwaway culture parallels the demise of our manufacturing sector? After all, people who don’t know how to make things will soon lose the skill to repair those things. Of course it could be argued that even poor Americans are too affluent to bother with repairs. Whatever the reason, we now have a consumer society almost completely divorced from production. Everybody knows where our widgets come from—China—but almost no one knows, or seems to care, how they get made. We fetishize the latest gadgets and venerate the entrepreneurs who run the companies who make them, but the self-evident miracle of creation itself, the process of designing and engineering and molding and assembling all our gear, takes place off camera. Whatever happens before the UPS truck delivers the new gadget to our door is like the era B.C. (Before Consumerism), a dark time when molten plastic roiled upon the Earth and our merchandise was formless and void. It’s as if the subject itself makes us uncomfortable, reminding us of the loss of our esoteric knowledge of the assembly line. Modern manufacturing is indeed complex, but the basic process can be readily grasped.
It begins, like most modern things, with a computer—in this case a program used by the designer to generate a detailed three-dimensional model of the product. This computer drawing then goes to a tooling engineer, who is charged with designing the best way to make the “tool,” which is factory lingo for the mold—a two-piece apparatus held together in a powerful vise and into which molten plastic is injected. The tooling engineer must figure out where on the mold to inject the plastic and how the mold will come apart after the plastic cools. Some shapes (like a sphere composed of two halves) are simple to tool. But if the product has a complicated surface with lips and returns, the tool could require elaborate moving parts called slides and pulls, which shape the plastic within the mold. That kind of tooling gets expensive; a good tooling engineer will always be looking for the most cost-effective solution, which may involve molding several parts for later assembly. Once the tooling engineer has designed the tool (or tools), a final 3D computer drawing of the part is created. That design is then electronically transferred to a milling machine, where a high-speed cutting wheel similar to a router bit automatically carves a full-sized model of the part out of a piece of solid copper.
This copper model, called an electrode, is then placed under heavy pressure against a block of steel and charged with massive electrical power at high voltage. The surging power and high pressure (tempered by a special cooling liquid) causes the copper electrode to physically cut a “negative” version of the shape into the steel, which becomes the actual mold. The finished steel mold is then fitted to a machine that injects it with heated liquid plastic that has been tinted with dye pellets to a specified color.
Elsewhere in the factory, workers (and robots) put together the product’s internal subassemblies—the circuit boards, LED boards, and other components that will go into the finished piece. Once the molded plastic pieces and the subassemblies are finished, the product goes to final assembly, where workers (in China mostly women) in white gloves and hairnets sit along a line, fishing the various parts out of plastic tubs.
But before thousands of identical widgets start rolling off the assembly line, the factory has to turn out just one very important widget—the final engineering prototype to be approved by the customer.
Whit slowly cut away the packing material on the DHL package from Shenzhen. “Here it is,” he said, “the final engineering prototype of the Burro light. There’s the Burro logo and the Greenlight Planet logo. Do More is, I guess, kind of legible. The color looks nice; they really pumped up the green. The battery lid doesn’t close as well as I’d like already; it’s not flush. Supposedly it’s keyed so it will only fit one way. Aha! It is keyed. I wish it would fit more flush. Let’s get some batteries and try it, shall we?”
Alec, the intern, grabbed three fresh Burro batteries. “Here goes,” said Whit, and he pressed the switch. The white plastic dome glowed with a dim but observable light, even in the daylight of the office. “That’s what we’re calling Super-Saver,” he said. “More than two hundred hours of light for about forty cents operating cost.”
He hit the switch again. “This is Saver—comparable to kerosene at one fifth the cost.”
Again. “Bright—much better than kerosene, still more than twenty-five percent cheaper.”
And again. “Super-Bright—more expensive than kerosene but the brightest light in the village. It’ll really light up your room.” He turned the light toward his eyes and squinted. “Wow, that’s bright!”
So was my brother’s face. I remembered what Beatrice, the Chinese rep from Whit’s first potential partner, had said in defense of her crappy headlamp: “Why Africa has more strict standard?” Finally Whit had a light that met his strict standards for Africa. “I’m thinking we could put kerosene out of business,” he said.
5. Sacrifice
“Don’t even think about it.”
“Think about what?” I asked, glaring at my brother. We were at the fetish market in Accra, where voodoo priests buy curatives and offerings like frankincense, dried herbs, monkey heads, and crocodile skins. It was probably my last trip to Ghana and I wanted to check this legendary place off my to-do list. Tucked into the bowels of the timber market outside the vast slum of Agbobloshie (known locally as Sodom and Gomorrah), the fetish market was definitely not on the tourist maps, and I ended up paying a guy a few cedis just to lead us there.
We probably could have followed our noses; the place reeked of rancid flesh and pungent, unidentifiable spices. Sellers glared at us from behind baskets filled with decomposing dog heads, their lips frozen in gnashing snarls. Desiccated, co
iled lizards were stacked artfully on tables next to crude wrought-iron daggers. Whole python skins, rolled up like sleeping bags and tied with raffia, shared table space with black, shrunken monkey heads and empty tortoise shells. Hanging from a stall like coats at Macy’s were dozens of whole animal skins—antelopes, crocodiles, and several large wild cats, the latter priced at three hundred cedis each (about two hundred dollars) but negotiable. That’s a lot of money in Ghana, which probably explains the cat knockoffs on the same rack—antelope hides on which feline spots had been obviously painted. The real cat furs were stunning and sad, luxuriously soft and improbably tame. “I said don’t even think about it,” Whit repeated as I brushed a feline pelt.
“What do you take me for, anyway?”
“I take you for someone who if he thought he could get away with it would smuggle a leopard skin home, or at least a leopard-skin hat.”
“Give me a break. These animals are endangered. I’d go to jail. And Sarah would divorce me. Is it legal to bring back a python skin?”
“It’s illegal to bring a python skin into my car.”
“What a wenis. But just as well. Hey, check out these voodoo dolls. This one’s painted white; it looks like you.”
“Seen enough? This place creeps me out.”
Voodoo wasn’t the only bad medicine bothering Whit. His giardia was still hanging on, and the doctor at the clinic in Koforidua had prescribed the nuclear option: a broad-spectrum antibiotic called chloramphenicol. While I was driving to Accra, Whit swallowed his pill and googled chloramphenicol on his smart phone. He read aloud: “‘Due to resistance and safety concerns, chloramphenicol is no longer a first-line agent for any’—repeat, any—‘indication in developed nations. In low-income countries, chloramphenicol is still widely used because it is inexpensive and widely available.’”
“In other words,” I said, “it’s being dumped on the poor, like all the other shitty products over here.”
Max Alexander Page 38