Sphero: from Concept Robot to Polycarbonate I - The Ball [VIDEO]
09 Oct 2012
Two years ago, Sphero was just an idea. Orbotix was nothing but a pair of young founders, Ian Bernstein and Adam Wilson, hacking on robots, trying to brainstorm something fun. Fast-forward to a passing grade from TechStars plus three rounds of funding and Orbotix was off to the races. We haven’t looked back. Just how much can happen in two years? Read on to find out in this first installment of the Sphero: from Concept Robot to Polycarbonate series.
Orbotix is now a growing company of 35 employees, all focused on making the most amazing robot in the world – Sphero. Glowing in the palm of your hand, Sphero seems simple. But it takes each and every one of our skilled team members to make a robot as complex as Sphero as intuitive as it is today. Our goal remains the same as the original idea of our co-founders: to make a gaming system that is fun for anyone who picks it up, something that can become many different things; a ball that you can race around, that can be used for tabletop games, that can be used as a controller, that can be programmed, and that continues to get better. We have some pretty lofty goals!
When we first started, we couldn’t control the ball. It wouldn’t even roll in a straight line. Our co-founder, Ian, and myself (CEO, Paul Berberian) were so depressed one day in November 2010 that we decided to pull out of CES 2011. The Consumer Electronics Show in Las Vegas was to be Sphero’s big debut, but without a drivable ball, it would be silly to attend. We gathered our handful of employees and made the announcement that we’d have to miss the event. And then something magical happened. Dave, our Lead Engineer, said “No, don’t pull out, I can make it work.” In my 20 years of running tech companies, never before has an engineer stepped up and committed to something that even Ian and I weren't willing to risk (and I take a lot of risks). But Dave delivered and got Sphero rolling. CES 2011 was a huge success for our team.
The event wasn’t easy. Sphero was still massively under construction. In fact, we had our junior developer, Skylar, stuffed in a 2-foot by 4-foot room for 4 days straight, repairing units as we broke them on the floor. At this point, Sphero was held together with rubber bands and needed to be taken apart to re-charge. It was only a rough draft of what we planned to build. And more importantly, we still had to figure out how to make it all work from an engineering standpoint.
For example, we told the world Sphero would charge by induction, but we had no clue how to make that work. As if making the ball roll wasn’t hard enough, we weren’t even sure if we could re-charge it! Opening the shell would destroy the experience altogether, making Sphero less maneuverable, less durable, and certainly not waterproof. I remember in March thinking, “Oh man. This ball idea isn’t going to work.”
Charging wasn’t our only challenge within Sphero’s first year. We also ran into these issues:
- The shell seemed impossible to manufacture because it had to be perfectly round, super thin, and super strong – three things that don’t go well together. We had no clue what the ball would look like until late Fall 2011 – literally just two months before launching.
- We couldn’t find motors that would last as long as we needed. Apparently the average toy motor is only spec’d for five hours (ours is over 150).
- Designing the robot to survive over 30 drops onto concrete from one meter high wasn’t exactly a piece of cake.
- We changed processors and had to rewrite all of our firmware.
- The gyroscope we used was so new to the market that obtaining parts was a nightmare. We had to order parts six months before production.
- We had no choice but to design the packaging months in advance, before we even knew what Sphero would look like.
- Completing FCC and other certifications in such a short time frame was a challenge.
ESD stands for electrostatic discharge, and it represents one of our biggest challenges in the development of Sphero. ESD is what happens when you get a shock if you rub your feet on the carpet and then touch metal. It’s annoying, but even more so for Sphero. Two weeks before production started, we got our final engineering samples. At this point, everything was supposed to be perfect. And in China, everything was. But when Sphero rolled around on nylon carpet in the cold, dry mountains of Colorado, ESD set its electronics haywire.
Because Sphero is 100% sealed, there was no way to ground out the static it develops rolling on the carpet. Instead, the static discharged onto the main circuit board. The solution was a last minute fix that worked up to 30,000 volts (we soldered a wire from the crystal to the ground plane and put heat shrink tubing on the antenna).
We missed on two fronts that first year. First, we underestimated how expensive Sphero would be to build. We also missed on timing. No one had told us that it normally takes two years to go from a piece of paper to shipping an original consumer electronic device with complex electromechanics. We did it in 16 months.
Since we launched in December 2011, we have shipped tens of thousands of Spheros worldwide. We have launched in major retailers including Apple stores, Brookstone, Target, Amazon.com, Future Shop and a dozen others. We built 15 apps on two major platforms (iOS and Android), hosted hack events for hundreds of developers, and had the joy of seeing people share their experiences with Sphero. And while this all sounds like a lot of challenges for a startup, the real story is in the software.
Stay posted to learn about the next installment of Sphero’s history – the software.