Sent email to WITHOUT THE x: markwxsherman AT yahoo DOT com
Please make sure there is an appropriate subject line.

Mark Sherman

Click here to view my other projects.


Click to view a movie in Google Video

Here is a second movie

Research Objective

The purpose of this project was to mimic the locomotive movements achieved by biological snakes. Snake locomotion is very interesting because it does not require the use of limbs such as legs or arms to move. Although many robots use wheels for locomotion, they usually become caught on obstacles or find areas inaccessible due to their size. To eliminate the use of wheels that limit the terrain that can be covered, some robots have attempted to mimic the walking motion of animals. Walking robots can range from very simple, to very complex, but often have trouble with balance or obstacles, and their movements are not usually smooth. The wave-like motion of a snake is promising for robotic development because it allows smoother locomotion which might become an asset if a robot was being used to carry hazardous materials, needed to survey in uncertain terrain such as the debris of a downed building after an earthquake or was needed to quietly and with low profile assist law enforcement personnel in survellance manouvers.

Another asset to a snake-like design is the low frontal presentation and elongated body which can wind its way through narrow openings such as pipes, breaks in walls, fissures and the like. A snake can move around obstacles and fit through small cracks and could be used to retrieve audio/video images from inside an area inaccessible by humans. If a robot were to mimic the movement of a snake, it would be able to negotiate through such small openings and there would not be any issues with balance due to its low profile. These aspects are important because in unstable environments, a machine must be robust enough and hardy enough to endure the treacherous conditions it is sent into, but also not contribute to the destruction or further obstruction of the target of the search. Although Slither is a work in progress, it has already realized many of the objectives of this project.


During locomotion, the servos move in the pattern of a sine wave. The underside of the robot has wheels, which have little friction forward and back, but much more friction from left to right. As the segments move in a sine pattern, the robot is forced to move forward, since each segment can only move forward or backward. The wheels resist side to side motion.

During locomotion, the servos move in the pattern of a sine wave.


Backwards locomotion is achieved by reversing the direction of the wave.


Turning left or right is accomplished by adding a curve to the sine,

producing a sine wave that curves in one direction.


Slither in full arc


Slither has also been featured in the RSA Robotics Expo at Fort Mason in San Francisco, twice, once in July 2003, and once that previous winter.

In addition, it was also at the 2001 SFRSA robot games at the Exploratorium.



Slither is a semi-autonomous robot. The whiskers (developed after pictures were produced) allow the robot to turn away from obstacles. When triggered by touch, a microswitch sends a signal and the robot alters its path. The robot can also be controlled more directly by the user with a 38 Khz TV remote control (or other, see below) which is picked up by the onboard infrared receiver. If the robot receives a signal from the remote control, it will cycle through various modes of operation with each press: forward, left, right, reverse, and frozen. The whisker sensors have the highest priority, and will circumvent remote control if the user tries to command the robot to run into an obstacle.


Schematic for Remote Control


More detailed specifications, including the program code, is contained within this paper. (PDF format)

10 segments made of bent aluminum

9 Tower Hobby Servos

Servo refresh rate of 13 Hz

Approximately 1 amp of current (with servos running)

servos run on 4 NiMH AA batteries.

BasicX 24 Microcontroller

Microcontroller runs on a seperate 9 volt NiMH battery.

2 whiskers, each made from a microswitch with a wire tie attached

infrared receiver to accept commands from any 38 Khz TV remote

Slither Schematic Diagram

The diagram below is the "brain" of the robot.


Assembly Notes

During assembly, a wire network resembling a spinal cord was created with wires of various appropriate lengths terminating at the servos. These wire bundles were channeled through holes drilled into the 10 aluminum framework segments.


New Objectives

Some considerations for future developments on this project include finding alternative ways to keep the smooth movement without the use of wheels, provide more sensory input to allow the robot to sense heat, and light. Slither should also be developed to be more independent. Audio and video delivery can be explored as well as experimenting with ways to traverse inclines effectively.

(c) Copyright 2001-2003 Mark W. Sherman, all rights reserved.