Re-inventing the Wheel

November 2011
By Bryan Bergeron

The field of robotics is so vast – spanning mechanical, electrical, and computer engineering – that you have to focus on one area to get anything done. For example, if you’re into pattern recognition-based navigation, you’ll have your hands full with image processing algorithms and perhaps image acquisition hardware design.

That said, it pays to occasionally broaden your focus and take a look at what’s happening outside the world of robotics. There’s a lot of activity in the electronics industry that can be applied directly to robotics, saving you time and money.

For example, I was looking for a way to quickly illuminate a scene with IR, alternating with visible light in order to improve a vision-based navigation system. I started with two banks of LEDs and an Arduino Mega processor. However, a few weeks later — while searching for LEDs on eBay — I happened across several inexpensive LED light arrays, all compatible with DMX512 controllers. Up to that point, I hadn’t heard of DXM.

It turns out that the DXM (Digital Multiplex) standard has been around for decades, and that it’s used to control dimmers, fog machines, and moving lights. You can check out the standard on the Web, but the point is that I was just about to replicate a small part of the DXM standard, unaware that I could buy — off the shelf — exactly what I needed. Compact controllers with an assortment of sliding and rotating potentiometers and buttons are readily available.

More importantly, for my project I located an inexpensive Arduino-compatible DXM board. The microcontroller board enables me to control the IR and visible LED banks through a USB link with my laptop, as well as run the banks autonomously.

Are you in the process of unknowingly reinventing a protocol or algorithm that’s been around for years? There’s something to be said for what can be learned by tackling a tough software or hardware problem. The problem is that time is limited, and you’re probably better off buying infrastructure devices and focusing on your strengths and interests.

Also, when you’re trying to solve a problem, the solution isn’t always selling on eBay or Amazon for a few dollars. It may be the problem and resulting solution were addressed in the past, with little or no commercial footprint. Take the hemispherical omnidirectional gimbaled wheel, or HOG wheel. It’s been around for nearly a century, but unless you follow RoboGames you probably don’t know how it works, or how you can apply it to robotics.

According to IEEE Spectrum (spectrum.iee.org/blog/automation) Curtis Boirum, a graduate student at Bradley University demonstrated a drive system based on the HOG wheel at the 2011 RoboGames symposium. The drive is a rubber hemisphere that rotates about the vertical axis, with servos that can tilt the axis left and right and forward and backwards. You can see two versions of the robot in action by searching for “Hemispherical Gimbaled Wheel Drive System” on YouTube. The speed and agility of the robot are amazing — there’s no way a human could evade this robot.

So, how can you keep track of both the past and the increasingly rapid rate of current innovation? I set aside one day a month to catch up on the old and the new. I’m fortunate to have several engineering libraries in my area, with books on past inventions. When I can’t make it to a library, I search the patent database at USPTO.gov. For current innovations, it’s hard to beat Google and — of course — contemporary magazines such as SERVO. SV