Total Manufacturing Assurance; Chapter 5, System Definition (page 4)
AUTOMATION
          Automation, from the Greek words for self and moving, is one of the most common buzz-words in manufacturing today.  In this section, we will look at some of the basic concepts of automation, and examine some of the strengths, limitations, and tradeoffs involved.
          The word automation can be applied to all parts of the manufacturing enterprise where direct human control has been minimized or even eliminated.  This includes material handling, machining, assembly, inspection, storage and retrieval, and any other tasks that are carried out in the process of converting raw materials to a useful and salable product.
          Automation is almost universally accepted as a good thing.  The reasons for this acceptance are compelling.  Productivity increases are an obvious advantage of automation, both in terms of production per hour and production per worker.  Other advantages of the automation trend are decreased unit cost, increased safety, increased product quality, reduction of work-in-process, and greater predictability of the manufacturing function. 
          Basically, automation can be broken down into two broad classifications: hard and soft.  Hard automation is the older of the types, and is based on machines which do only one thing, but do it extremely well.  The auto industry was dominated by hard automation a decade ago, and still makes extensive use of it.  Examples of hard automation include automatic presses and die-casting machines, automated transfer lines, and any other heavy machinery that performs the same task over and over.  It is characterized by a lack of flexibility, long production runs, and high production rates.
          Soft automation, on the other hand, is characterized by machines which can perform a variety of tasks.  Soft automation is also called flexible or programmable automation, and is dominated by robots and other general purpose, programmable devices.  The idea behind flexible automation is that a single investment in equipment will pay dividends in the production of many different products and processes.  One of the costs of this flexibility is that the tasks are sometimes performed at a lower production rate than if dedicated, single-purpose machines were used.
          Which type of automation is best for a given application?  The answer to this question depends on many factors, including the type of industry, the anticipated production rate and run length, frequency of product changes, and future business plan.  For any one given manufacturing task, it is much cheaper to acquire hard automation equipment that can perform the task than soft automation equipment. 
          However, if the task becomes unnecessary, the hard automation equipment will require extensive modification, if not outright scrapping.  The flexible equipment can be easily modified for a new task via software alone, with no further investment in hardware.  Furthermore, the flexible equipment can be instructed to perform the new task along with the old task, as each one is needed, if a variable product mix is desired.
          As a final note on this topic, it should be mentioned that the distinction between hard and soft automation is not as clear-cut as it is often made out to be.  In reality, a sort of continuum exists, where any level of flexibility that is desired can be achieved.  There are individual devices that exhibit certain degrees of flexibility, and there are entire factories that contain elements of the programmable along with the inflexible.            The amount of flexibility that is selected for any specific installation should be based on the amount that is needed and anticipated.  To pay for excess flexibility that will never be utilized would be an unwise waste of funds.
          
Automation Elements
          All automation is based on certain types of equipment, capable of performing standard automation tasks such as material handling, assembly, inspection and so on.  Each of these, with the exception of robotic devices, are available in varying degrees of programmability, and as such should be considered key elements of both hard and soft automation systems.  Robots and similar devices, of course, are inherently flexible, and should be considered elements of soft automation only.
          Transfer mechanisms move a workpiece from one station to the next in an automated fashion.  Several configuration exist.  An in-line transfer mechanism consists of a conveyor or overhead-chain system to move the workpiece along in a more or less straight line, while a rotary mechanism is more like a turning table, around which the necessary processing stations are arranged like guests at a dinner table.  The rotary configuration is useful only for situations with a small number of workstations, whereas an in-line system can accommodate as many stations as are desired by merely increasing its length.  Both types can move in either a continuous or indexed fashion, depending on the types of operations to be performed at each station.
          Automated machining devices perform the same functions as traditional machine tools, but are directed automatically with some sort of computer control, generally CNC or DNC (discussed below).  These machine tools can consist of traditional tools such as drill presses, vertical or horizontal milling machines, lathes, presses, and the like, or more modern devices which combine several of these functions.  The machines may operate upon a stationary workpiece positioned by the transfer mechanism, or may have a built-in multi-axis positioning table to hold the work.
          With regard to the newer multi-purpose machine tools, many types and configurations are available.  These are often called machining centers, and may have several spindles driven by a common motor, and capable of coming to play on the workpiece in whatever order is required by the process plan.  A typical machining center may combine a drill press, vertical and horizontal mill, and tool holder.
          PLC, CNC and DNC refer to methods of utilizing computers to control automated machine tools.  They will be discussed here in order of increasing sophistication.  However, it should be kept in mind that new hardware and techniques are constantly being developed, with the result that clearly drawn distinctions between these three types of control are becoming more difficult to make.
          PLC stands for programmable logic controller, the original device used to control automated machine tools.  A PLC is a dedicated device attached to one tool, and often is supplied by the machine tool manufacturer as a part of the tool system.  It is basically a digital computer which can store a sequence of operations for the tool, and send the appropriate commands to the tool at the appropriate times.  The input console of the PLC often consists of a collection of buttons corresponding the standard commands for the tool.  A small screen is often included so that the program can be reviewed and edited.
          The PLC bears little resemblance to a general purpose computer, but is more of an intelligent command console with a memory.  Some are equipped with disk drives so that programs may be saved, and some have serial ports allowing them to download programs from some central computer.  PLCs have been popular since the late 1960s, and continue be important industrial components today.
          The DNC concept moves toward the use of general purpose computers, rather than specially designed digital devices, to control machining processes.  In DNC, or direct numerical control, a central computer stores the programs for a large number of machines.  It can be thought of as one large PLC, managing all the machines in a factory or part of a factory.  This allows for coordination of machines and processes, as well as monitoring of overall factory activities.
          CNC, or computer numerical control, was a natural outgrowth of DNC which was made possible by the decrease in size and increase in power of general purpose computers.  A CNC computer only controls one machine, or possibly a small number of machines working together.  The advantages of CNC over DNC arise from the localization of control.  An engineer on the factory floor has direct access to the computer at the site of the tool.  Also, since CNC computers control a single machine, there is no need for time sharing or other resource allocations, and the complete power of the processor can be utilized in complex control algorithms.
          There is no danger of a single computer failure crippling the entire factory, as in the DNC implementation.  CNC computers are usually networked to a central computer, so that there is no loss of the DNC ability to monitor all processes and keep all elements of the factory in communication with each other element.
          Robots, or more precisely, programmable manipulators, are the final element of manufacturing which help to pull everything together.  Robots have many applications in an automated factory, including loading and unloading pallets, applying sealants and adhesives, cutting, inspection, and assembly.  Perhaps the most popular use of robots currently are painting and welding.  Any task which involves the manipulation of some material or device through a precisely defined trajectory can be accomplished with some type of robotic device.
          As with any programmable automation element, a robot must be able to justify its high price tag through savings in manufacturing cost.  Tasks which are performed many times per hour, perhaps for two or three shifts per day, are especially deserving of robotic attention.  Also, tasks performed under hazardous conditions, requiring many expensive protections and precautions for human operation, may be much cheaper for a robot.
          Several issues must be kept in mind when considering a robot for a task, based on the robot's capabilities and the task requirements.  Chief among these is accuracy: can the robot position and orient the object to be manipulated with the required accuracy?  Related to accuracy is repeatability.  Repeatability is often defined as a robot's ability to reach the same location several times in a row, with out drifting as time goes by.  Another issue is resolution, the smallest increment of motion possible for the robot.  Is it small enough to perform the most delicate of the required motions?
          Other, more practical considerations include such things as load carrying capacity, reach of the arm, and type of grippers (also called end effectors) available for the robot.  Another is the capability of the robots controller; how programmable it is, and what kind of sensors it can be connected to.  These capabilities vary widely from one type of robot to another, and the options available on the market today are quite vast.