How would you describe your first encounters with hardware?
My interest in hardware was sparked when I was eight years old and my dad bought an Apple II clone. Back in the day, it came without a case, so all of the electronics were exposed. I could see the electronics and I wanted to fiddle with it. My dad didn’t want me to touch it because I might break it, but when he wasn’t home I would still fiddle with it. I broke it several times because the chips are in sockets; even though my dad told me not to, I just wanted to see what happened when you put them in backwards. I learned very early on that putting chips in backwards is a bad thing!
The great thing is that Apple II came with a cool set of schematics and source code. I was the weird kid in elementary school who carried around an Apple II reference manual. On the playground, I would just pull up the schematic and stare at it because it was so fascinating. I didn’t understand what I was looking at, but I had some inkling about the connection between lines on the schematic and wires on the board. Over time, I learned to map those to functions bit by bit and piece by piece, and it all started coming together.
How did your early experiences affect your decision to go into the hardware industry?
I just kept snowballing from there and learned more. I went to MIT and I flipped a coin and instead of going into biology I went into electronics. I got a degree there and eventually went to industry, hated that, then went back for my PhD because I wanted to hide in my shell a little more. After getting my PhD, I participated in a bunch of startups that all failed. I never had a successful startup, but I learned a lot from failure.
I did some silicon chip design, some reverse engineering, and then eventually got to the phase in my life where I started doing manufacturing. For many years, I was looking to do the biggest, baddest, hardest project I could do, which meant working for a pure tech startup. With something like that, you’re way in the future and basically by the time the technology works and goes onto the market the patents have expired. There is no capital monetization and you work really hard and it’s really obscure. As a result, I never had anything ship in volume. That was the most frustrating part: to put your life into something and never have it see the light of day.
What lessons did you learn while working on Chumby?
I got tired of working for a pure tech company, and so decided it was time to join a company that could monetize a business idea quickly. So, when Chumby came along, I joined that and I wanted to do open hardware and manufacturing and started logging up my experience in manufacturing and hardware. We worked and produced the first Chumby and then multiple generations after that.
I was working on that project from 2005 to 2010. When I started Chumby I had never mass-produced a product or done mechanical design, and I didn’t even know what injection molding was. But I had the privilege of sitting with some of the other engineers at PCH, and I would just get on the factory floor, see what they are doing, and learn about it. Eventually by the time I was through with Chumby, I was able to use SolidWorks to design my own cases and do injection molded cases from scratch.
After this, we started talking about what the next project would be. We did a series of projects such as reverse engineering some SD cards, and I met Jie Qi and then I helped her to produce Circuit Stickers under the “chibitronics” brand name.
At the time, one of the guys working with me in Singapore was Sean Cross, and him and I were sitting around asking what we should build. We decided to build something we can use, because when I was at Chumby I built things for other people, rather than myself. I use a laptop every day and we needed a development platform, so we built the laptop that we would actually use. We are now doing a crowdfunding campaign around that product.
“I never had a successful startup, but I learned a lot from failure.”
How would you describe your process of going from a prototype to manufacturing it?
There is actually a whole lot of art in manufacturing and designing things from the ground up to be easy to make. One of the great things is to have an approach where you are fully responsible for your own supply chain. I don’t like to have a supply chain manager and a manufacturing manager. I want to make something myself. I insist on doing all of the testing myself. I insist on handling the manufacturing issues myself because, from the design standpoint, is it forces you to think: can I build that? If I gloss over this bit of detail, I might pay dearly for that later.
From the very beginning when you start designing, you are starting to think about how you make it manufacturable. What manufacturing process should I use? How do I make sure I can source all of these components? How do I make this all work? When you actually get to the manufacturing time, you have made all of the decisions because you are the one who has to pay the price at the end of the day. That’s generally been my approach.
What do people most overlook when they are designing?
There are a lot of things that you could forget. I think the two that come to mind are, firstly, the ability to source the materials, and secondly, the yield. For example, if you look at the instructions for a cool project in Make magazine, oftentimes it tells you to go find an obscure or out-of-date object, like a motor from a 1980s VHS player. In theory that would be great, but then all of sudden everyone is going to eBay trying to find the same part, and it’s not sourceable despite the fact that many people have this cheap item in their garage or are just throwing them away.
On the yield side of things, a lot of people won’t run the numbers in terms of what it means to be yielding, and so every step you have some fallout, yet every step is about 99% yield. If you take ten steps like that, your yield will then be at about 90%. You get people who essentially build the leaning towers of Pisa into their project and at the end of the day the problems compound, preventing delivery. It’s crucial to build a system that is robust and re-workable so that every step can be coupled with another step to minimize the yield fallouts. Otherwise, you are throwing away a lot of money.
How would you describe how things have changed in the perception of hardware since you got involved in manufacturing?
It’s weird because right around the time that I was doing the work on the Xbox in 2001, hardware was probably at the rock bottom. It was around the time of the dot-com boom. It was really super hot working on Web 2.0 and you had to do something with Amazon or XML and if you did that it was cool. Soldering was a low value thing that happened somewhere else. But I was still that weird guy who knew how to solder in a lab so people would come to me with broken things and I would fix them for all the time. I just stuck with it because that is what I do, and I love doing it. One of the reasons the Xbox security was relatively easy to break was because they had this assumption that hardware was hard and that soldering was difficult. Actually, if you know how to do it, it is very easy. I did it on a grad school budget for about $150.
I gave some talks at conferences after the Xbox hacking, basically telling people that hardware is not hard — there’s no magic behind it. I showed people that the “magic” was actually pretty simple manufacturing techniques. Then all of a sudden Kickstarter came to be, and money started going into a system where it hadn’t before, because all of the VCs wouldn’t touch hardware; they thought hardware was a retail chasm where all this money had to be paid up front and basically the startups all die, and you don’t get returns.
So, all of sudden these cool companies began raking in a million dollars in Kickstarter as their seed round and delivering on their products most of the time, and there’s nothing like money to get the interest of the guys in Silicon Valley. Since that’s happened, we have had a radical change. People are starting to get into hardware more and more. The problem is a lot of people think they have to add hardware now, yet they have no idea how.
Another problem now is that there are increasingly a lot of scams on Kickstarter, where there are all these hardware bits and pieces and people can’t tell what’s real or what’s fake. I know the industry is definitely feeling like a bubble already. I can sense the bubble growing now I think maybe I liked it better when nobody knew about because at least I didn’t have to worry about competing with fraudsters.
“That was the most frustrating part: to put your life into something and it will never see the light of day.”
How have you approached finding your own factories?
I did a talk the other day at Generator and one thing I talked about was how to approach factories and how to engage in building relationships. If you’re a startup and the only value you can bring to a factory is money, then you are basically worthless. The startups don’t have any money and if you have money it’s a finite amount and all of the factories know this.
A lot of startups want to go to somewhere like Foxconn, but Foxconn has a ton of people and capability. They don’t need your help, but they do need your money and you don’t have a lot of it. If you try to engage with the really hip factories, you find yourself depleted of cash very quickly and you can’t launch.
I look for factories that are missing a little bit so that I can give them value that’s more than money. When I come in with my product, I will help train their staff to be capable of building my product. They see value in that and you get to that point where you are building this relationship where it’s more than money that you are trading.
What’s the challenge for online hardware startups when they get to the retail phase?
There is the world of the internet and online where everything is automated and it seems like you could solve any problem with technology. But retail is all about the salesperson meeting buyers face-to-face, doing demonstrations, and going to the Walmart or Target headquarters to actually develop relationships and cut deals. It feels like an older system and a lot of people don’t expect that because they are doing business with Kickstarter.
The problem is people want to physically see the product before they spend a couple hundred dollars on something. Best Buy is becoming a showroom for Amazon these days, but offering the product in-store is really valuable. If you are going to make that big of an investment you want to be able to see and touch and feel. There is probably room for some disruption. You can try your hardware yourself and describe it to people, but at the end of the day, retail presence is needed in order to sell hardware effectively. If they start a business online from the beginning, margins are much fatter online so they tend to under-price their products and then when they get to retail they can’t survive.
What are some of the most common questions that a hardware entrepreneur will ask you?
It really depends upon the background of the team. The questions they tend to ask usually center on what is weak in their team composition. Some guys have super hotshot electrical engineering guys. No problems there, but they have no mechanical engineering so they have a bunch of “mech-y” questions. Some teams will have essentially no electronic engineers at all and then the big question is how to create a hardware startup with no one who can design electronics.
Generally they tend to be technical teams so they are often weak on the marketing and business side. Some of them do have business guys involved early on who can map it all out and get a strategy in place, but a lot of them have great tech ideas and think they are going to launch and introduce this stuff without realizing they’re missing crucial aspects to their strategy. At that point, I get them to tell me what they’re doing and give feedback. It’s almost not the questions they ask, but what they forget to ask, which are the important things they need the most help with.
“It's almost not the questions they ask, but what they forget to ask, which are the important things they need the most help with”
What do you think is missing from the startups that will be necessary for the ongoing support of the hardware ecosystem?
There are a lot of things missing. There is a huge mismatch between the way manufacturing has been done and the way it needs to be done to match these more agile, lean, and honestly, less-experienced companies. I don’t think it’s in an impassable chasm. Basically the ODMs who have factories and resources need to raise the level of service. A lot of people expect the ODMs to be able to answer a lot of questions. There are unreasonable expectations between startups and manufacturing because they can offer absolutely zero insight into costing down your product. People get upset because they just don’t see that conflict of interest.
A lot of people say China should be cheaper. They don’t understand. A factory is not a designer; their job is to ensure that your board works. Typically if there is a yield issue or they substitute a part in that is cheaper, it’s not really their problem at the end of the day. Why would they find a cheaper part for you with full risk of that design change, and then give you the cost down and then earn less net margin? A lot of people get mad at factories for not being more aggressive on keeping the cost down, but if you think about it, you really have to get engaged and you need to get an engineer working with these guys to cost things down because ultimately it’s your bottom line. It’s your net profit. You don’t just go to China and let them do it.
An ODM can possibly solve that problem by getting people on staff that can do that, but then they charge for the service of the cost down, so it is now sustainable for the ODM to do that. It helps to have an increase in variety of services offered by certain platforms so at the end of the day, when the innovators are using these platforms, there isn’t any mismatch with what they are actually used to, versus what the startup actually needs.
It would be good if there were more interoperability in the industry. One of the startups I work with is Spark; they really try to enable people to use their hardware platform by being open. I feel like one of the pieces missing is they need to get ODMs to be “Spark certified.” A lot of times someone wants to design one product into another product, and the suggestions about how to do that effectively are all over the place. Even if you have all the necessary information, it’s not a streamlined process for most people.
At the point when someone is given all the design answers, there is still a lot of decoding that has to happen. Even the bigger companies are afraid of that because they don’t have the competency to hire the people to get that done.
What is your current focus in the hardware industry?
Right now I’m working with Jie Qi on Circuit Stickers. We are getting to the point of shipping the units out and I’m very hell bent on making sure that I meet the deadlines I set for my campaign. I actually want to ship on time and get things to people when I say I will get things to people because there has been way too much lateness in crowdfunded campaigns. It doesn’t have to be that way. You can do it on time, you just have to set expectations and you have to have your stuff together before you announce the date. You need to know when the inventory is pretty much there and ready to go. We have a number of lines that are selling and about half the lines are done with manufacturing, they are just waiting in the factory to ship. A couple of new ones are behind, but we still have until May to solve these issues. I think it will be no problem and so I am looking forward to seeing that line grow and develop and work with more people.
The other thing I am working on is this laptop project Novena, which we weren’t really planning on doing last year. This is the one I built for myself; I built this handmade prototype last December, which is this like little kind of crummy, leather and paper thing that I put together and we used it to give a presentation at CCC and the response was overwhelming. I thought that was great, but knew we have to make it manufacturable. I did a re-factor of the design to make it more manufacturable and more sourceable. The campaign seems to be going well so far. I think it will fund and I’m looking forward to getting it manufactured and out in the world.
What have you learned from your two crowdfunding campaigns?
Now having completed almost two crowdfunding campaigns, it’s given me a lot of insight into it. I had mentioned earlier how people selling online price their product too low to later move into retail, but for me it’s been really painful to now maintain the high price that I say that everyone else should maintain. It’s so tempting to go lower to an unsustainable point.
The reason a lot of crowdfunding campaigns fail to deliver is because they price too low — they can’t actually build it for the price they said. Even knowing this, I still had to grit my teeth on the laptop because I had to price it higher than I really would have liked it to be. Despite the high price, if we were to close the campaign at exactly the amount I hope to raise, I would probably just barely not lose money on it, but a lot of people don’t see that. If you look at something like the Ubuntu Edge, they raised $12 million but needed $25 million to succeed. That’s because in order to get to a price of $700-800 per phone, they had to build 40,000. So even though people thought it was cool and it raised a lot of money, it didn’t reach its funding goal, which is a sad conclusion for everyone.
When evaluating a price point, I could either price my laptop much lower and need thousands of people to buy to reach my goal, or I could service a really focused market of a few hundred open source enthusiasts who are totally on the same page as me. At the end of the day, especially in the early phases, you really want those people. They are going to be your best users. You want to take care of them and them give them best service possible. You’re going to charge a little more, but you’re going to build a really good product for them and they’re going to be happy. That’s a much happier conclusion in my mind than trying and failing to shoot the moon.
What are you excited about for 2014?
I take it day-by-day. I don’t like reading the news a lot and look for trends and events. I just do things. I am hoping that my products will ship on time. I am also looking forward to Burning Man; I love Burning Man. I might end up doing a project for Burning Man, but it usually ends up being one at the last minute. We will have to see how it balances with everything else because I do have these two campaigns going and if they really start taking a lot of time then I have to set my goals and priorities correctly.
In terms of a trend for 2014, Moore’s Law is slowing down. This is actually allowing small companies and startups to be more innovative. Before, if it took you a year and half to launch a product, something twice as good is out there; but now it’s slowing down and lowering costs.
When and where were you happiest?
Right here, right now.
What is your idea of misery?
A corporate job dealing with a product that I know will never go to market.
Which living person do you most admire?
My life partner.
What’s your favourite quote?
“It didn't work because it wasn't supposed to work.”
What book are you reading right now?
Stranger in a Strange Land by Robert A. Heinlein
If you were to die and come back as another person, who would it be?
Not sure, but if I’m reincarnated, I’d like to be female. Women are pretty awesome.
Which talent would you most like to have?
I wish I were more creative -- that I could be better at visual arts, or play the piano.
What is your greatest regret?
Dying without making the most out of the life that I have.
What do you consider the most over-rated virtue?
Justice is in the eye of the beholder.
What is your greatest extravagance?
Making my own laptop.
Randolph Wang from Intel Labs China asked, “Do you have any personal desires that you see yourself doing in five years?”
If I had a perfect world, I would be doing what I am doing today. I am really happy with the freedom that I have right now. When I started my “unemployment phase,” I changed my priorities.
When I first started out from college, my original hope was I would get rich quick in a startup so that I could then take the time to do what I want to do. That’s not going to happen; I failed at enough startups. I’ve come to terms with the fact that I’m not going to get rich quick — that’s just not happening.
So instead, I decided that first, I am going to have fun, and second I am going to do something good for society and third, it must be sustainable, but just so. That formula all has brought me to a point of happiness in my life.
Ideally I can continue doing something like this, like designing another generation of laptop. I hope I can continue to help people with interesting ideas and teach them. I really like teaching and helping people. I hope I can continue to just live a healthy, happy life.