First Steps. Why a Robot Sim Startup?

When I began this startup company, savvy business people told me that I need to figure out "who the customer is" and "what the customer needs". They said "You are out to relieve some sort of pain or provide a service or product that people didn't even know they need". While this may be on the mark, it misses the one key ingredient: Obsession. Not just passion (I really want to do this), but obsession (I really need to do this). I suspect that most startups are about the need of the founder(s) rather than the need of the customer. To do something that they've always needed to do. And maybe that's what customers want most of all: people to be obsessed with the products they provide.

My obsession with robots and engineering began when I was 11 years old. My parents took me to the Ontario Science Center - a vast complex of technology and engineering built as a museum. I loved that place the way some kids love a baseball field or soccer pitch. In the museum, someone had built a crude cyborg leg made of wood and four pneumatic (air) pistons. The pistons worked as the muscles, expanding and contracting. Two pistons for the upper leg to move the hip and two for the lower leg to move the knee. I spent a long time controlling the pistons to move the leg which was attached to a bike pedal. Round and round she went, once I figured out the pattern of activating the pistons. These patterns of motion were to become the seeds of the Xemo project years later.

When I think of robots, I think of humanoid cyborgs - two arms, two legs, a torso and a head. Two or four-wheeled carts with sonar sensors do not really do it for me when thinking of robots. Braitenberg's book on Vehicles is really cool, but doesn't embody the "robot" concept for me.

Robot locomotion on two and four legs present a real challenge to human engineering and ingenuity. If you think robots move quite well today, then compare them to human movement. Go to the ballet and watch the grace and smoothness of the dancers. Look at the amazing balance of a mountain biker tearing down steep terrain over rocks and dirt. And observe the spinning and flipping of high divers, gymnasts and circus acrobats. Today’s robots are no way near this level of dexterity, movement and grace.

I realized that I couldn't create the Xemo robot sim game alone. I've gathered a team of experts in robotics, evolutionary algorithms, physics, game design, programming and art design. With the Xemo game we want to let every robot fan experience the fun of getting a robot to balance, stand up, walk, jump, crawl, rollover and about a hundred other actions. How many motions can a robot perform without sensory feedback? How much better when we add sensors? How can basic robot movements, called behaviors, be combined into a complex set of movements to navigate through an obstacle course where the obstacles move as well?

For me, robots are about integrating hardware and software to execute physical tasks that humans find so easy to do but are so hard for us to program. We're building Xemo so that people can create these complex motions in simulation.

How amazing would it be if we could get millions of kids (and kids at heart) to experience the fun of getting their very own virtual robot to take its first steps?

We're off and running now!