Introducing OAP: An Open Reference Design for Building Intelligent Mobile Robots

"Have you ever dreamed of building your own Linux‑based droid that could roam around your home autonomously, intelligently obeying your commands?" asks Dafydd Walters. "You may now finally bring that dream to life." Walters founded and leads the Open Automaton Project (OAP), which empowers hobbyists to assemble an intelligent mobile robot equipped with stereo vision and a cutting‑edge PC‑grade mainboard—all for the price of a high‑performance personal computer.

While the RETF and OROCOS initiatives focus on defining open standards for robot hardware‑software interfaces, OAP takes the implementation route. Its goal is to produce a reference design that bridges the gap between sophisticated research robots and the small, rug‑roving models typically built by enthusiasts.

The project’s SourceForge-hosted website offers circuit schematics, source code, and comprehensive documentation under an open‑source license, enabling anyone to build their own autonomous mobile robot.

Key project objectives include:

• Designing a coherent set of modular components—both hardware and software—that adhere to industry standards wherever possible.
• Keeping the total cost between $1,500 and $2,000, roughly the price of a premium PC.
• Developing a low‑cost real‑time vision system that serves as the primary spatial sensor.

To stay within hobbyists’ budgets, OAP relies on readily available consumer‑grade hardware. For example, the vision system uses off‑the‑shelf FireWire webcams mounted on a pan‑and‑tilt head driven by standard radio‑control hobby servos.

Leveraging the recent proliferation of consumer‑grade webcams and FireWire‑capable motherboards, the project has made low‑cost stereo vision achievable. The chosen camera is the ADS Pyro 1394 WebCam, though any IIDC‑compliant (DCAM) digital camera will work.

Augmenting the vision array are twelve sonar detectors (Devantech SRF04s) managed by the oap‑sonar command‑line utility. One unit is mounted on the pan‑tilt head to mimic eye movement, eight are positioned around the base to detect obstacles, and three face downward to sense ledges.

An infrared sensor, tuned to the human body’s emissivity, is mounted on the head and helps the robot recognize people.

The pan‑tilt head is controlled via an I2C interface and the oap‑head utility. Mobility is provided by two main DC drive motors, managed by oap‑motor. Wheel encoders supply odometry, while a power‑management module monitors battery levels and can communicate with a docking/recharging station via infrared.

OAP’s human‑robot interface (HRI) includes:

• Speech output
• A minimalist robot‑mounted LCD and keypad
• Wireless and infrared remote controls

Speech recognition input will not be attempted initially, according to Walters.

For more details, see the article: Meet OAP — an open robot reference design project.