Roboty's Status Page, Taygeta Scientific Inc.
The current status of Roboty:
- Had some terrible mechanical problems at the Exploratorium Robotics
exhibition (broke 3 legs !). Now doing some mechanical upgrades to fix
the problems. The basic problem is that the robot has a rather narrow
domain of optimum mehanical stability and the sensor head has changed the
balance and weight enough to be significant.
- Started work on an articulated "sensor head". The head can turn left-right,
and tilt up-down, it is driven by a pair of servomotors left over from the work
on the leg design. This platform contains: the feeler bump sensors,
the IR sensor, a pair of Polaroid
sonar sensors, a visible light detector and a tilt sensor. At this point I am
moving the pre-exisiting sensors to the head and have not yet integrated the
new sensors into the robot.
- I finally got a reliable IR proximity detector working. The
IR sensor is designed to provide object detection just beyond the range of
the feelers (around 30-40 cm). There are two multiplexed IR LEDs, one on
each side of the head, and a single IR detector in the middle so that the
robot can detect obstacles on either side.
- I brought Roboty to the San Francisco Robotics Society
Robot Sumo contest
at San Jose State
(as a demonstration robot, it does not qualify as a competitor). During the demo,
Roboty walked across the stage and stepped up into the ring on its own! It is
believed that this is the first robot to have done this.
- Added a pair "feelers" that provide contact bump sensors. These are
just micro switches attached to some very stiff wire which protrudes ahead of
the insect on either side of the "head".
- Once the machine could walk on its own and use an onboard supervisor
CPU, I found that while the original nibble wide "bus" was convienient for attaching
to a PC parallel port, it was rather demanding on my I/O pin budget. So I
the Motorola SPI for synchonous serial communication on the 6805s.
This resulted in a drop from 9 to 3 I/O lines necessary to control all six
- I am doing initial tests with using a
New Micros 68332 board as the supervisor. This board is not
as convienient to work with as the Basic Stamp-II that I originally used,
but it has vastly more computing horsepower.
- I have done some "field tests" and have demonstrated that the robot has
the strength and stability to be able to walk on moderately rough terrain
(e.g. a typical suburban front lawn).
- The robot is walking on its own! It can walk forward and backward with
several different gaits, (all the stable
periodic gaits described by Song and Waldron).
- I am fine tuning the gaits used for turning
(the balance is different).
- I have developed a set of data structures that make it easy to write
down a gait and then implement it by simply properly initializing the
- We have cut the umbilical !
The robot is now walking with its own supervisory CPU onboard. I am
using a Basic Stamp-II from
Parallax as the supervisor. I have found the Basic Stamps to be a
delight to use. For power, it is carrying a 6V 3A-Hr battery, which appears to
be capable of running the robot for several hours (I have yet to significantly
drain the battery).
So far no sensors are being used, so it is walking blind.
- I am now adding sensors so that it can be truely autonomous.
Back to Robotics page.
Skip's home page
Everett (Skip) Carter Phone: 831-641-0645 FAX: 831-394-5561
Taygeta Scientific Inc. INTERNET: email@example.com
1340 Munras Ave., Suite 314 UUCP: ...!uunet!taygeta!skip
Monterey, CA. 93940 WWW: http://www.taygeta.com