Christian Carlberg is a Cornell-educated mechanical engineer. After his undergraduate degree, he spent three years doing classical aerospace work and getting an MS, part time, at night school. At 25 he moved to LA to work in movie FX, specifically “practical effects”—physical, animatronic robot-based puppets now in decline thanks to low-cost CGI. After that, he moved on to show-ride engineering for Disney, where, among other projects, he worked to make the 6-foot diameter giant squid eyeball at Tokyo Disneyland’s 20,000 Leagues Under the Sea attraction light up. He left Disney to become a regular and successful contestant on the TV show BattleBots.
These days Christian works in commercial robotics at Palo Alto’s Suitable Technologies, developing a sophisticated remote telepresence system called Beam. Since 2011, in his free time, he’s been leading a small team (two initially, now expanded to six) to design, develop, manufacture, and distribute the world’s first fully open-source, user-programmable high-performance LED flashlight.
Christian’s a busy man, by anyone’s standard, but I got a chance to catch up with him last month and get his insights about what it’s like to successfully crowd-fund a hardware development project with a very small team.
You originally launched the HexBright project at Maker Faire, right?
Correct. I made the Bay Area Maker Faire 2011 my target date to launch HexBright on Kickstarter. I enlisted the help of my friends to get a website together along with video and pictures. I literally filmed my promo video the day before Maker Faire and launched on Kickstarter the first day of the Faire. I spent both days at the Faire handing out fliers. Most people were interested, but more than a few people were skeptical. That’s OK, I’ve been called crazy before.
Your Kickstarter was hugely successful.
When funding closed, eight weeks after the Faire, HexBright FLEX had raised almost $260,000, making it one of the top ten most-funded Kickstarter projects at the time. I had more than 3,000 backers. Mythbusters‘ Grant Imahara gave me a video endorsement!
You started with the idea that you wanted to do a Kickstarter campaign, and then decided an open-source flashlight was the right project for that purpose. Why?
Three reasons: One, I had some LED experience, from my time at Disney Imagineering. Two, a flashlight is an interesting blend of multiple disciplines—mechanics, electronics, battery management, optics, as well as aluminum and plastics manufacturing. Three, I had an interesting idea about how to seal the micro-USB port.
Tell me about that, the USB seal.
You will note that the inside carrier threads are at the front, which draws the rear o-ring to seal against the body. I thought it was a great way to have a rechargeable flashlight with very good sealing. It’s a four-start thread, by the way; there are four separate grooves which overall makes it very coarse. But it seals tight and in one revolution you have access to the micro USB plug.
Why did you decide on open source?
I looked around and saw that no one had really done it. I realized I had a niche market.
What qualities of the flashlight are programmable?
You can think of it as having three inputs—the rear button, the little black reset button on the board, and the accelerometer—and three outputs: the main lamp (which is a CREE XM-L U2, the best available right now), a rear green indicator LED, and and a rear red indicator LED.
The light includes an accelerometer? What for?
Whatever the heck you want! Part of the idea was that I wouldn’t program anything but let “the wild” figure out stuff. I supply the tool, you supply the creativity. People are putting programs on Git Hub.
But at some point, during the design process, there must’ve been a moment when you said, hey, let’s put an accelerometer in the thing…
Actually, that idea came from the backers themselves, in the very early days. Sure, it costs more to put the part on the board, and I don’t use it at all in the factory-installed program. But it’s there if you want to hack your light.
So the design evolved during the funding cycle.
I originally offered two designs: a simple, small, cheap flashlight with a cool grip design, and a big, more expensive, open-source flashlight, thinking most people would want the smaller one. I was waaaaaay wrong. Everyone wanted the cool open source design.
You had a custom aluminum extrusion made for the body. Had you done custom extrusions before? How did you find the manufacturer for that?
Yes, I’d done it before, but it’s actually pretty easy. You just make a drawing and a .dxf file of the cross section, and send it to a shop for feedback and a quote. There are a ton of shops in LA and all across the country that can do it.
I recently saw a quote from the developer of the videogame FTL, which was a huge Kickstarter success, to the effect that he probably wouldn’t do it again because crowdfunding adds “a whole new layer of stress.” What do you think?
Yes, actually, that is true. There is a world of difference between having a finished product to sell and trying to design a project in front of thousands of eyeballs and potential complainers. You can’t please everyone all the time, so someone always complains. But right now all my US backers have a light and they love it. I always knew I could pull this off and make a quality product, but it took time to get there, and I had to deal with a lot of people constantly asking “where’s my flashlight?” It’s kind of like driving a busload of kids who won’t stop asking “are we there yet?”
What was the most fun part of the process for you?
The very start and the very end! The original design process, and shipping the units. Everything in between was a grind. I had to deal with suppliers, manufacturing issues, costs, Kickstarter backers, and general logistics. All very critical and challenging but not terribly creative. It was a fantastic experience for me, but every light that went out was a small load off my back.
How would you do it differently next time?
Offer fewer options, and ask for more money per reward than you think, but still slightly less than “retail.” Try to get more homework done ahead of time. I used Kickstarter as it was originally intended—as a platform for project development and not so much as a store. As a result, I literally started with a rough idea and developed it all the way through to a product. That takes a long time and a lot of effort, especially when you’re also holding down a day job, as I was. I am pretty pleased that in my spare time I created a quality product in the same ballpark as big, multi-million dollar name brand flashlights.
HexBright FLEX is available for purchase now at HexBright.com.
MAKE Volume 34: Join the robot uprising! As MAKE's Volume 34 makes clear, there’s never been a better time to delve into robotics, whether you’re a tinkerer or a more serious explorer. With the powerful tools and expertise now available, the next great leap in robot evolution is just as likely to come from your garage as a research lab. The current issue of MAKE will get you started. Explore robot prototyping systems, ride along with the inventors of the OpenROV submersible, and learn how you can 3D-print your own cutting-edge humanoid robot for half the price. Plus, build a coffee-can Arduino robot, a lip balm linear actuator, a smartphone servo controller, and much more