2014-04-28

Co.Labs

Tough Questions You Face When Building A Homebrew Wearables Kit

Here's an expert run-down of the sensors that make a wearable device useful--plus what they do and how much they cost.



When the duo behind MbientLab tried to make a consumer product, it ended up where a lot of billion-dollar ideas do: on Kickstarter. “It failed miserably," says CTO Matt Baker. "What we learned after many months is building a brand and a consumer product is challenging. Instead, we decided to go back and stick to our guns as engineers,” he says. What they came up with next was timed perfectly for a world that is just waking up to the killer potential of wearable computing.

Instead, they built the MetaWear board: a wearable electronics homebrew kit that allows anyone to create a production-ready wearable device prototype "in 30 minutes or less." No programming skills are required. It works out of the box. And it lets people play in Bluetooth-connected products who previously wouldn’t have had the technical chops. (And yes, this one is on Kickstarter too.)

The Unexpected Challenges Of Building An Unbuilt Kit

The founders were faced with two big problems: how to decide exactly what to include in the kit, and how to source those components from the most reliable, economical suppliers. If you haven't looked--I hadn't--the number of sensors, buzzers, and batteries on the market is staggering. There are a number of places to source any given components. And with new wearables for tracking, safety, alerts, and identity popping up all the time, new types and iterations of sensors are ubiquitous.

“We took about six months to figure out what to put on the board," says CEO Laura Kassovic. "There’s only so much space you have to work with. And a two-dollar-a-pop humidity sensor significantly eats into the margins. We had to figure out: What sensors do 90% of the people want and why?”

Image Courtesy of MbientLab

The pair soon realized that solving this spec would require as much social engineering as engineering of firmware and hardware. It involved meeting with all sorts of customers and competitors; creating detailed spreadsheets of possible sensors, chips and batteries; finding the best overseas vendors; and figuring how to work in China.

“I learned, but a few months ago I really didn’t know how to do a proper purchase order," says Kassovic. "Business skills matter."

“We started by taking 50 or so meetings,” says Kassovic. That led to the creation of extensive spreadsheets of possible sensors, chips, batteries, trade-offs, and pros and cons of including each. It was eventually whittled down to a short list of must-have components any hardware manufacturer should consider.

Click to expand

The Must-Have Wearable Components

“The first two things that we really knew everyone had in common were the accelerometers and RGB LED," says Kassovic. "We wanted our device to be wearable. Right now there’s a lot of fitness trackers, and they all use accelerometer technology so it was a no-brainer for us to put that in. They’re really cheap too and you can do lots of cool things. You can detect 3-D motion--XYZ motion and things like free-fall. So that one sensor has a lot of applications. Everyone wanted that.”

It was the same for the RGB LED, she explains, because LEDs are the only way to show user feedback. "So we put the cheapest but most luminescent light we could find and we put it right in the middle of our board," she says.

Next, they added a button, "following a discussion with a couple of jewelry designers who were doing a Life Alert type of jewelry," she says. "They noticed that it feels very natural to grab something--the act of pushing a button is very natural--[when] you’re in a life critical situation." So a hardware button became a "no-brainer" too.

The other thing that a lot of people wanted was the coin vibrator--you know, the thing that buzzes inside your phone when it's on silent. But as Kassovic explains, this one required trade-offs. "The vibration from a coin motor was the perfect option," she says, "but, well, it’s huge, right? We went a little further to say, What is something that makes the trade-off between having this coin vibrator and having a smaller board?" In the end the team gave people the option of putting on the vibrator motors themselves. It comes as part of the kit, but it’s not soldered onto the board.

The gyroscope is a good example of something they decided to ditch. "We found that the applications for gyroscopes are really really complicated, like if you are actually going to do very accurate 3-D provisioning measurement algorithms get really complicated," says Kassovic. "And we didn’t want to make a complicated product, so we just left out the gyroscope entirely. Plus they were more expensive than the accelerometers."

Image Courtesy of MbientLab

Procurement 101: How To Get The Best Prices And Where

Ultimately, the components that make the grade for wearables are the ones with the broadest applications, smallest size, and the lowest cost. But for the MbientLab team, sourcing was also important.

"If I use the temperature sensor that was made by only one company somewhere in the middle of Taiwan, then I would get in a lot of trouble if they decided to stop making it tomorrow," says Kassovic. "You want to get parts that are readily available, which means they should be made by more than one company. Relying on on one company is generally a poor choice."

So where did they go to find their hardware and how did they go about finding it? “When we source parts for our board like the accelerometer and LED we first do a BOM--bill of materials analysis. This gives us a rough estimate of what the components on the board will cost and it helps us understand what the retail price of the board will be,” says Kassovic.

“Generally a lot of people when they’re first looking at component prices, they’ll go on the major websites like DigiKey and Mouser, and you’ll get pretty good approximations. And from there we can only expect the price to go down," says Kassovic. "The best thing you can do is to find the components in China, and you’ll probably get the message ‘this price possible’ and you usually either have something already in your repertoire or you already have the Chinese contact. Or you can start by going on alibaba.com and at least get some names through there, and then you start contacting and calling."

"You try and get the best price possible," says Kassovic, "but you also have to understand that you’re going to get a good price if you put in a bulk orders. For example, for our project, our order is 5,000. No matter what, we order at least 5,000 pieces of everything. But it decreases the price quite a bit too.”

Image Courtesy of MbientLab

Lessons Learned From Their First Failure

This team's tips for any aspiring Kickstarter startup: develop cultural skills, think beyond Kickstarter, and plan to scale from the beginning.

  1. Have Your Buyers Lined Up:
    “The first lessons we learned: A lot of people create products and they don’t know how to sell it. We were those people,” says Kassovic.“We thought we had a great product, we engineered the crap out of it, we were so proud. And we didn’t have a way to sell it. It certainly wasn’t Kickstarter," she says. "That showed us everything about what not to do. Before we even created it we should have had either somebody like Walgreens or online like Amazon.com or Think Geek saying ‘Yes, we’ll sell it.’ But we didn’t have that. You need to already have somebody who is willing to pay for you to make 1,000 [of your product]. Some sort of customer who says ‘Yes, I’m going to put them on my shelf and I’m paying for 100 to 1,000 of them right now.'"
  2. Develop Cultural Skills:
    “Matt’s the hardware engineer. I’m the social engineer,” says Kassovic. “I send 200 emails a day just to get pricing on batteries and things like that. A lot of people don’t understand that’s part of the development process. For us it’s really easy to create the CAD. But you also have to go and befriend the nice guy from China and then maintain the relationship.” According to Baker, “it’s really culturally dependent. It depends which country you’re going to source from. Some countries expect more interpersonal relationships. Some countries, like the U.S., it’s more transactional, it’s just a business deal. Even a basic culture book will kind of explain what the business culture is like. China, for example, is the kind of place where you want to get to know the people working in the factory on a personal level," he says. "Because if you reach out to people lower down on the chain who actually do the day-to-day work, when they know you, they’re very friendly and very open to telling you about issues.”
  3. Think About Scaling From Day 1:
    There is an old British saying--"start as you mean to go on." Startups often think about the early stage of development, but the most successful ones also think like the big shots. They think about how to scale the business and what will be necessary to do that from the very beginning. “Time and investment in relationships can actually be more important at scale," says Baker. "If you eventually hit the 100K quantity mark and you’re talking about a lot bigger numbers. An email once in awhile or even Skype... goes a long way,” he says. Kassovic adds, “Once you get to a point where you’re making things and ordering a quantity of above 5,000, actually it’s better for you just to go to frickin’ China. Go see the manufacturers and see what they’re actually producing. Usually when you’re creating something in the tens of thousands, you’re just going to have to travel.”

[Image: Flickr user Benjamin Chun]




Add New Comment

0 Comments