Robotics - A Mechanical Extension of Man's Ability
My earliest thoughts on robots originated from those cheesy early sci-fi movies and television programs of the 1950s. As a child, I would on occasion have nightmares of robots chasing me, slow humanoid-looking bots that I was always able to initially outrun, but robots that never stopped coming for me. Yes, they were slow, lumbering, but relentless! While in junior high, I realized that I could also build robotic devices that would at least roll around and look ominous. After all, I knew about motors and I was pretty handy with axels, fan belts, batteries, and pulleys. My first serious attempt to build a robot came when I was 16 years old. By that time I had learned to use an arc welder and an oxygen-acetylene torch. I cut and welded a frame with 1.25” by 0.25” metal bars. I mounted a winch motor someone had given me, along with a 6-volt car battery purchased at Western Auto for $8.00 plus tax. With a windshield wiper motor that I reengineered to be reversible, I created a fairly sophisticated steering system. After modifying the main drive motor so that I could change the field polarity of the motor, I had a robot that could move both forward and backward. Using only one rear drive wheel with two steerable wheels in front, I eliminated the need for a differential.
The holy grail of robotics, both then and now, is total wireless control. In the mid-1960s there were very few ways of controlling such a robot available to amateurs. One method was through the use of reed relays. This relay has a set of 5 tuning fork-style reeds each of different length, all driven by a single coil. When you drove the relay with a certain frequency, one of the reeds would resonate and start an intermittent closure each time the reed hit an electrical contact. I purchased one such relay and started to build a unit that would generate 5 different audio frequencies that could be used to control 5 different channels in an on-off manner.
I had some problems with crosstalk using the 5 oscillators, and at that time did not understand electronic bypassing as I do now. Although the controller did work to an extent, I never saw the full realization of the system. I only used the robot via an umbilical cord-type connection to a switch box for control.
In 1978, I decided to build a robot modeled after my earlier high school attempt. I did not know exactly how I was to control the device through a wireless link, but I had decided to build the frame and worry about how to get control signals to the bot later. I went to a local supplier and purchased some 1-inch square steel stock that I could cut and weld in my garage. Since I did not have an act welding machine at that time, I had no facilities for anything but torch welding.
I cut and welded the tubing to form my basic frame, and even went so far as to purchase chain sprockets for driving the two main wheels. The third wheel was a coaster-style wheel at the front. After welding the frame and planning out much of the drive system, I moved back to north Louisiana, and placed the partially completed frame in my storage room.
In 1994 while working at LSUS, I decided to complete my robot. I finished and painted the frame and went on to complete the electronic control system which consisted of a 220 MHz amateur band transmitter and receiver. I sent TV signals back to the control location using a 470 MHz ATV link, also amateur-based equipment. Yes, I have a ham license if case the reader is wondering.
I named my creation, “Roving Observational Vehicular Autonomous Robot” or just ROVAR. I was suddenly a very popular invite to a variety of functions. Actually, it was not so much me, but ROVAR that people wanted to see. One of my students at the time, Mark Mains, was taking a triple major in Physics, Computer Science, and Math. Mark could absolutely think on his feet and was the best ROVAR operator that I had ever used, and that includes me. He was a master at selecting the actions and phrases for ROVAR to speak when confronting newbie ears.
After years of demonstrations and exhibitions, I informally retired ROVAR. I was in a way happy that I could tell people he had been decommissioned, or placed in moth balls, and could not come out to play. Finally, however, in the mid 2000s I decided to donate ROVAR to LSUS so that I could use parts and equipment owned by the State to revamp my creation. I turned ROVAR into a student research project and had my students start working to redesign and rebuild the aging bot. As time went on more and more unique innovations were added. Now, ROVAR has reached a far more sophisticated level of entertainment for those wishing to have him visit. He is also seen rolling down the halls of the science building and speaking to people as he will.
In the early 2000s, I was designated as the robotics teacher for the Computer Science course CSC 450 pertaining to the programming of robot arms using standard industry interfaces and languages. One semester I decided to teach this course from a different perspective and after teaching the standard material for about a half of the course, I allowed my students to have a group activity to design a mobile robot.
We were seeking a robot that could negotiate rough terrain, something with a track or at least multiple wheels. The students and I designed Wheeled Hybrid Electronically Engineered Linear Motion Apparatus or WHEELMA for short. WHEELMA was a mass undertaking that both excited my students about the possibilities of engineering and computer science, but also brought out their creative side. WHEELMA took a couple of years of work to realize the final product, but “she” is now a mainstay at in the Science Building and has interacted with many students as well as outside people.
My publishing the research aspects of WHEELMA helped to obtain the needed work to obtain my rank as full professor. So what is so unique about WHEELMA? Well, having a full machine shop at LSUS is a tremendous asset to both faculty and students. Using this facility, we were able to create a robot that is supported by 4 separate wheel sets. Each of these wheel sets consists of two wheels driven by a common drive axil. This drive axel is center in a hollow shaft that allows the rotation of the wheel sets. In other words, a set of two wheels can rotate, or “float” as the bot negotiates upcoming terrain. As the wheel sets rotate to conform to the terrain, they are driven for propulsion by a center shaft that is independent of the rotation of the wheel sets (See photo albums).
WHEELMA works very well, and although the wheel sets do not fully rotate as was initially planned, they do rotate back and forth enough to take upcoming terrain or steps. The rotation is driven by 2-inch bore pneumatic cylinders under electronic control. Amazingly, this is all that is necessary to take steps. Once the lead wheel is elevated and in rotation, it is easy to take upcoming steps.
Future Robotic Work
A number of years ago, I spent a considerable amount of time thinking about an aerial robotic platform that could be electronically stabilized using rate gyros and integration of their signals. After giving much thought to the approaches possible, I settled on a concept of four electric motor-propeller combinations that could in total produce enough lift to levitate the robot vertically. Rather than vanes or control surfaces, I decided that you could have two motors rotate clockwise and two to rotate counterclockwise. By differential control of motor sets one could both level and control rotation of the robot. In other words just by RPM changes in the motors, one could establish 3-axis control.
Using students, we worked on this project to master the required technology. Due to a variable number of interested students, this project has been off and on for several years. During that time, I have seen my concept used in a multitude of flying robotic craft. Termed a “quadrotor,” they are even being sold commercially, with numerous universities using them in unmanned drone applications.
I have several more ideas for variations on the flying robotic (drone) designs. One set of ideas uses a gasoline engine for central propulsion. The heavy lifting can be done with a brute force engine, while smaller variable speed electric motors can be used for stability and stearing. I also got an idea from Georgia Tech's robotic competitions from back in the 90s. I am not sure now which university built the model, but the design has merit, but the control system was not what was needed to make it function. This craft used one engine and 8 control vanes for stability and navigation.
David Poe, one of my student research students, is shown on the left with our quadrotor project. David put a great amount of time into perfecting the control system of this flying platform. David has graduated from our physics program and went on to obtain a MS in electrical engineering From North Texas State University in Denton, Texas. David was at one point a political science major at Louisiana Tech before coming to LSUS and starting his study of Applied Physics. He was a good sport about all the claims that he was capable of returning from the Dark Side of the Force.