Home Chapter 13 AUTONOMY-HBridge Motor Control with Pulse Width Modulation and an H-Bridge Motor Controller
AUTONOMY-HBridge Motor Control with Pulse Width Modulation and an H-Bridge Motor Controller
“Only a machine can appreciate a sonnet written by another machine” ~Alan Turing


Question: How can I control the direction and speed of two DC motors?

The neurophysiologist Dr. Grey Walter (1910--1977) explored theta and delta waves using an electroencephalogram (EEG) that amplified brain waves produced by the brain in light and deep sleep. In the 1940s, Walter did early research on autonomous robots at the Burden Neurological Institute in the process of modeling human brain function. He studied reflex behaviors to test his theory that complex behavior can arise from the interconnections of neural circuitry. In this research, he created "Elsie" and "Elmer", two robot "tortoises" that showed very complex behaviors in navigating their environments using a phototube eye and two amplifiers equivalent to two neurons. These transistors that drove relays that activated the steering and driving of the motors. [1]


Replica of Walter Grey Turtle built by the "Intelligent Autonomous Systems laboratory, University of West England, Bristol" from Dr Alan Winfield. 2008.


Walter’s research was influential in the birth of the field of cybernetics, which was documented in his book "The Living Brain" (1953). Grey’s research would influence researchers like Rodney Brooks of MIT, who observed that insects with relatively few neurons could display very complex social behaviors and navigate their environments with great skill. Brooks was interested in using simple rules to create complex and emergent behaviors, which he termed “subsumption architecture”, and created a series of insect-like robots that are very successful in navigating complex environments. These systems are capable of taking the unkn own and responding intelligently


Subsumption architecture allowed a system to emerge into complex behavior while relying on a series of sensors running independently.

Emergent behavior results from the interaction of an agent within a dynamic environment. Here, behavioral patterns are not programmed in, but rather result from, the system components interacting with the environment [2]


These sensor networks would trigger various subsuming behaviors like attraction and repulsion. For example, if a limit switch were not triggered, then the robot could move forward based on a heat sensor. Therefore, we could say the limit switch subsumed the heat sensor.

The writer Valentino Braitenberg describes a series of thought experiments in the book Vehicles: Experiments in Synthetic Psychology. Braitenberg describes how simple, internal structures can give rise to unexpected and complex behavior, and he speculates that humans might ascribe love, aggression, foresight, and even optimism to these vehicles because they don’t realize the underlying principles that make them behave in the way they do. In other words, meaning is opportunistic.

Braitenberg sees this as proof of what he calls the "law of uphill analysis and downhill invention,” which states that it is more challenging to guess the internal structures necessary to generate complex robot behaviors than it is to create structures that can actually exhibit such behaviors. [Ibid]

What if these vehicles moved across the body instead of the floor? The artists/inventors Erwin Driessens and Maria Verstappen created an autonomous robot called Tickle (2000--04). Tickle is a small robot that rolls on the body to generate pleasant, tickling sensations. It has two motors that power tiny rubber feet for a good grip on the skin. When it senses a slope that is too steep, it will steer away until a safe level is found.


Tickle by Maria Verstappen and Erwin Driessens. 2000--2004.


Tickle’s behavior is hardwired to avoid steep edges, and therefore, the body as an environment becomes a kind of dynamic programming for the work shaped by the Individual bodies. This allows Tickle to manifest different behaviors in relation the body this is exploring. Two high-torque, 5vokt mini motors drive tread that are silicone rubber caterpillar tracks designed to stick to the body. [3]

Petit Mal (1992--95), by the artist/theorist Simon Penny, is an autonomous robot that explores architectural as well as human space as it rolls through its environment on two bicycle wheels. Penny used multiple sensors and programming that activated motors attached to the bicycle wheels to create a sense that intelligence is present, though the intelligence is neither “anthropomorphic nor zoomorphic”. [4]

Penny thinks of this intelligence as unique to its electronic and physical nature. In Penny’s words, “An autonomous robotic artwork marks out a territory quite novel with respect to traditional artistic endeavors as we have no canon of autonomous interactive esthetics.”


Petit Mal by Simon Penny. 1992-1995.Centre George Pompidou.


In this work, Penny used microcontrollers, programming, an H-bridge (an integrated circuit that controls the speed and direction of motors) with two high-torque DC motors, and ultrasonic sensors for sensing walls and humans to help control and move this reactive and interactive work.

The overall system emerges with its architectural and human environment, and permits autonomous, sensory-based navigation.

Penny envisioned the Petit Mal system as an embodied intelligence, where the programming that is not prespecified is allowed to emerge from the bottom-up as the hardware interacts with its environment. Accelerometers are employed as minimal sensory input and still allow complex behavioral manifestations.


Petit Mal by Simon Penny. 1992-1995.Centre George Pompidou.


Accelerometers are used for measuring the effects of acceleration or gravitational forces. They come in many varieties and many animals have biological systems have cellular structures with pendulum like weights that can measure and give the neural circuitry a sense of gravitation pull or acceleration.