Affordable Robotics Platforms

My first exposure to hobby robotics was an inexpensive preassembled robot. Because programming the carpet roamer was both limited and tedious, within a few weeks I followed my initial purchase with a kit — this one based on a BASIC Stamp with ample support for programming. Eventually, I made the move to develop robots of my own hardware and software designs. My first creations were ungainly, ugly affairs. However, with time and experience, both appearance and performance improved.

If you stay involved in robotics long enough, you’ll eventually feel compelled to design your own beast. And doing so needn’t be an expensive, resource intensive endeavor, as long as you shop and design judiciously. One of the many advantages of beginning your experience in robotics with a programmable kit or an assembled robot is the availability of software. For example, let’s say you develop a program for the Parallax BoeBot, equipped with the line follower accessory kit. Once you’ve debugged your PBASIC code for the BASIC Stamp at the heart of the BoeBot, you can post your routines to online user groups, and everyone with a similarly configured BoeBot should be able to use the code. Similarly, you can immediately use routines developed and posted by others using the same platform. In this way, you can accelerate your learning and leverage the debugging experience of others.

When you develop a robot of your own design, you’ll learn something about mechanical assembly, strength of materials, power tools, and, in general, how to function in a metal working shop. On the downside, unless your design is destined for mass distribution, you’ll be largely on your own in terms of sharing code and adding mechanical peripherals. Lacking standardization between platforms, every custom robot is unique, even if based on a standard microprocessor configuration. Differences in, for example, the basic physical configuration of the platform, placement of wheels and type of drive mechanism, total mass, number and placement of sensors, and nature of peripherals prevents easy sharing of code and routines.

Is the relative isolation and lack of a general support community a concern? Of course. But the educational rewards inherent in designing and building your own robot outweigh the potential negatives. One of these negatives, prominent in the current economy, is the cost for going custom. It’s easy to spend several times the price of a full robot kit on a bare-bones platform. For example, one major robotic supply house sells 14” aluminum discs for nearly $50 each. A few discs and spacers can easily approach $150 for a bare-bones platform.

One way to minimize custom platform costs is to repurpose non-robotic platforms. My current favorite is the two-tiered Kik-step rolling step stool (Cramer Industries, [url=http://www.cramerinc.com]http://www.cramerinc.com[/url]), about $50 from Amazon and other online dealers. I picked up a stainless model, but they’re available in 10 colors. Regardless of the finish, the round step stool is stable and virtually indestructible. The three supporting braces between the two steps provide ample protection to any electronics mounted to the top of the first step, and the spacious, hollow underside is perfect for a pair of drive motors and wheels and battery pack.

I mounted a CrustCrawler Smart Arm ([url=http://www.crustcrawler.com]http://www.crustcrawler.com[/url]) on the top of the upper step, with a 7.2V NiMH battery mounted to the underside of the top step. I mounted ultrasound sensors from Parallax and an Atmel processor card on the larger first step, sharing the battery with the Smart Arm. The spacious underside encloses a pair of six inch wheels driven by Parallax 12V servo motors ([url=http://www.parallax.com]http://www.parallax.com[/url]) and two, 7.2V NiMH battery packs.

The Kik-step isn’t the perfect mid-sized robot platform. For one, steel is heavier and a little harder to work with than aluminum. The second is that, at an elevation of 14 inches, my CrustCrawler Smart Arm can’t interact with objects at floor level. These modest limitations aside, the price is right — especially if your goal is creating a swarm of mid-sized robots capable of autonomously playing bumper-cars.

The Kik-step is just one example of how you can save significant money by repurposing non-robotic hardware for your next robot platform. If you’ve developed a robot using an inexpensive, readily available platform, please send me a note so that I can share your experience with other readers.  SV