Automation, a term that came from the auto industry, was initially used to describe relatively inflexible mechanized systems that controlled one or more machines and the means of conveying work pieces between the machines. Numerical control (NC) was a term used to describe the process of controlling a machine tool, originally using a tape or punch cards like the old computers. Today's automated equipment uses state of the art computer technology for automation and machine control. Technology and automation advancements are evolving rapidly and the modern industrial plant reflects the evolution with ever more complex systems and ever more difficult challenges to provide an acceptable level of safety.
Human beings tend to want to be in control of machinery and not let the machinery control them. Human - machine interactions are highly variable and are affected by personal factors such as familiarity with the technology, uncertainty about processes and procedures, comfort with computers, knowledge of the systems, culture, language, age, education level, innate patience, and experience. Interactions are also driven by human factors engineering. Information display characteristics such as symbology, labeling, character readability, location, conformance with stereotypical expectations, and consistency with characteristics of input systems can affect the efficiency of normal production operations but might also have disastrous implications when a system malfunction is occurring. Similarly, human factors deficiencies with input devices such as levers, switches, buttons, knobs, pointing devices, can be problematic as well. The proliferation of complex and rapidly changing systems coupled with the factors affecting human machine interaction variability are a breeding ground for catastrophic injury. The frequency and severity of injuries is often related to the degree to which we fail to recognize human variability and take steps to control it.
Technology-driven systems are subject to the same learning curve problems as any human interaction. (Ott and Campbell, 1979.) Error rates tend to be high at first when the new technology is introduced and then decrease as we learn to operate the system. However, research appears to show that there is a finite minimum error rate. (Duffey and Saul, 2003.) Recognizing that there is a finite error rate and considering the amount of kinetic and potential energy in automated systems, the onus is on the designers, vendors, integrators and users to establish reliable safeguarding systems that are forgiving of error.
One of the major contributors to the automation evolution is the Programmable Logic Controller (PLC). Hardwired or wireless, these units are becoming ubiquitous. A recent change in NFPA 79 allowing safety PLCs having redundancy and reliability to replace hard wired safety relays creates both opportunities and threats for driving down risk on the production floor. Before this change, machine or automated system operation and machine safety were "wired" differently. State of the art computer systems would be used to provide the operational flexibility needed for quality and productivity whereas old style electromechanical systems would be used in parallel for safety functions.