Optimization
Introduction
The best things in life these days are usually expensive.
------------------------------------------------------------------------------
----------------------------------------------
"Is that the best we can do? There has to be a better way! There just has to be!"
This is a common refrain. We are often dissatisfied with the way we do business, with the way we deal with friends, with how we do our work, with the way we play the game. We are always looking for ways to do better. That's our objective here -- to find ways. And the following is a sample of how we might go about the task, conducting studies of ways using simulation, keeping in mind that a perfect model is only a dream, an ideal that's far out of our reach, except possibly in very special circumstances, with pretty much closed and static conditions. We simply don't have perfect "vision." So let's look at the example.
Imagine the following scenario:
You are seated at a life-size television screen, viewing a computer-generated replica of yourself serving a tennis ball. Watching yourself, you're not satisfied. Output from the computer says the service speed isn't competitive. Also, the output indicates the racket didn't quite engage the ball as required and you missed your target. So changes are needed. So you re-adjust parameter values, reset the starting conditions, and try again, not ecstatic but encouraged by the study.
This is an example of a personalized experiment. It aims to find the best way for you to strike the ball to deliver a specific serve. Similar programs can be constructed to study returns. I've taken a step in this project with a package of ten programs to study the racket mechanics problem, for both serves and returns. See Persona;ized Tennis Diagnostics.
I am currently studying the overall tennis problem of tracking, intercepting, and hitting your return shot, then moving to a best defensive position to receive your opponent's return shot. Needless to say, this is a complex batch of skills.
--------------------------------------------------------
Computer simulation is an excellent way to learn about dynamic processes, including bio-social behavior as well as physical movement. It has already been used to examine the motion of aircraft, spacecraft, missiles, and the planets, and the behavior of economic systems. Using computer simulation we can also:
Studies using computer models provide numerical data to help find optimum, or near-optimum, courses of action. They also help establish the relative sensitivity of variables to variations in other variables, so you can learn which are most sensitive to performance effectiveness and thus determine which aspects have to be given more attention in training and development. You can make small changes in the variables, one at a time, to see the effect they have on results. The benefits are useful in sports, for example. Click here for other examples and the analysis and design involved in such studies.
----------------------------------------------
Computer modeling is a normal part of design and development in aircraft manufacturing. Models are used to define flight characteristics of drawing-board creations and relate them one to another. Examples are maximum flying altitude, wing span, lift forces, maneuverability, rate of climb, and turning radius.
-----------------------------------------------
The spacecraft industry has produced simulated versions of astronauts maneuvering in gravity-free environments. Astronauts are depicted manipulating hand tools while assembling a space station, say, or repairing external antennae or solar cell arrays on a spaceship. Working without gravity has its problems. Make-believe helps to understand the actions and anticipate difficulties that might arise. Unfortunately, though, a supercomputer or banks of mainframe computers and many man-hours of programming are needed to produce acceptable models of events at such a high level of complexity and risk. The spacecraft are very complex, especially considering their redundancy and other safety features, But the flight personnel themselves are far more complicated than the vehicles.
---------------------------------------------
Simulation is used to study the effectiveness of production line designs. Picture a bottling factory, for instance, and visualize a line of stand-up empty bottles marching along a conveyor belt, passing a succession of worker robots, each performing a specific job. At one station a pair of metallic hands might reach out, grab a bottle and hold it firmly but not break it, as a second pair of hands moves a nozzle into place and fills the bottle with cola. The hands then, let's say, release their grip, and the bottle moves on to the next robot, whose job is to clamp on a snap-top cap. The bottle might finally be wiped dry and labeled and packaged, all without a single human hand touching it. A make-believe model of this bottling setup might reasonably be developed to study its feasibility.
---------------------------------------------
Researchers have adopted the computer to study the movement of athletes in order to improve equipment. This has enhanced the quality of rackets, golf clubs, skis, shoes, sleds, and skates, among other things. The ultimate aim is to optimize performance.
Other scientists are currently using models to study physical stress -- to uncover faults or find better ways to execute specific skills. In these studies math formulations define the event in question, and variation of starting values of the variables or values of the parameters yields alternative results. Here, again, aspects of both the individual and the environment are generally considered to be part of the model. Both internal and external forces are therefore examined.
In a simulation of this type I represented the motion of a tennis athlete on the court, visually tracking and running to intercept the ball. My goal was to depict the player's visuo-motor processes: to portray the athlete perceiving the ball and running to intercept and slam it back to his opponent.
--------------------------------------------
Whether you are flying an aircraft, building an orbiting space station, hitting a tennis ball, tracking a potential terrorist, or caring for victims of violence, you are a perceiving, moving emissary that makes decisions in one setting or another. You observe what is occurring in the environment, attempt to keep track of relevant items in it, and perform designated actions.
Sensory-Motor Interaction
Perception and movement are highly interactive. Movement is needed for perception, and perception is needed to have useful movement. We activate eye muscles to accommodate to the visual objects or to coordinate the joint action of our two eyes for binocular vision. In a similar way, fibers of the muscles have to be active in order to generate feedback signals to the brain, and we are constantly making decisions about what to view and how to do it. Movement is integral to our perception of the environment.
Everyday Phenomena
We continually use our senses to keep track of things and guide and control our actions. As our main source of information, we commonly depend on our eyes, but also use peripheral and internal receptors, though we may not always be aware of it.
In these circumstances, personal forces interact with social and physical forces; we operate in a complex field of forces. We can direct our own actions. But in order to move effectively we have to be aware of our intentions, external objects, and the forces exerted by the objects. Knowledge is required. In short, we have to know our turf. This is a complex and highly valuable tracking capability. Indeed, man is a perceptual being.
So we find ourselves playing tennis according to our own rules on special courts. We drive manufactured cars on streets of our own making. We fly aircraft in designated air space. We run power plants, work in bottling plants, buy groceries at the local market, wash dishes in special sinks, and on and on. It is behavioral environments of this kind that we consider when we apply the computer to build simulations for study.
-------------------------------------------