The objective of the project is the design and construction of a small autonomous low-cost six-legged walking robot, code-named ITSY. Emphasis for the project is placed on a well-documented solution to the goals outlined below.

• design and construction of a mechanical device to emulate biological 'walking', using six modular legs and an electronic control system
• design and utilization of subsumption architecture behavior modeling to formulate intelligent reactions to sensor input and an internal world-model of the surrounding environment
• documentation of the entire process in a report and webpage available to other students and robotics enthusiasts:
  • (http://will.sitch.org/projects/itsy/)

• design and construction of a 'payload' of electrical equipment to fully utilize the mobile platform and the advantages of legged locomotion

Most mobile robotics projects consist of payload structures supported by differential-drive wheeled systems. The wheels are easy to control and power, and supply a high weight allowance for sophisticated payloads. Design and construction of wheel-based systems is relatively simple and highly documented.

However, there are a number of drawbacks associated with using wheel-based systems as a platform for other tasks, most notably rugged terrain. Tracked and wheeled robots have trouble if the height of an object it must overcome approaches the radius of the wheels. Additional problems occur if loose dirt or gravel provides slippage beneath wheels.

One solution to overcoming complex terrain, while providing a stable and rugged platform for further electronics equipment, is legged locomotion. Using a collection of twelve servomotors (two motors per leg), each leg will be allowed two degrees of freedom, enabling several different movement gaits and speeds of travel. This particular solution is not unique, but is a complicated exercise in hardware and software design.

Using modular hardware devices (designing each leg as a building block) and building a subsumption-like behavior through object-oriented programming practices yields a package that can easily be improved or enhanced with additional work or peripherals.

• design and construction of a walking robot to overcome complex terrain while retaining the ability to deliver an electronic payload
• use of biological 'walking' patterns to allow for an efficient method of machine-powered locomotion
• documentation of the design and construction, to be used in future research projects as a basis for continued study

ITSY will be approximately 40cm long, 25cm wide, stand 20cm in the 'raised' position, 12cm in the 'lowered' position, and weigh less than 2 kilograms. The control system will be based around the MIT HandyBoard, using the Motorola MC68HC11 processor, and two Mini Serial Servo Controllers (Mini SSC-II) designed by Scott Edwards Electronics. The software will be loosely based around subsumption architecture theory initially proposed by R. Brooks.

The modular leg design is the result of almost six months of analysis and brainstorming. The final design used with ITSY is loosely based on many existing MIT robots, including Attila and Ghengis. The leg design uses a mechanical system to optimize servomotor torque in specific regions, while 'locking' when the leg is extended and bearing the weight of the robot.

ITSY will use a large array of sensors to build an internal 'world-model' of the surrounding environment, from which the software will make assumptions and switch between different behaviors. Subsumption behavior theory outlines the complex layers of 'behaviors', and the stimuli required to move between layers of operation.