Automated Control Systems Do They Reduce Human Error And Incidents

WITH EACH PASSING DAY, the world becomes more technologically advanced. As automation, artificial intelligence and robotics improve, it may be increasingly tempting to employ automatic means to accomplish production goals. Just think of it: If no humans were on the production floor, no more human errors would occur, no one would get injured, and companies would produce higher-quality product that they could move faster to market. Supervisors would not have to deal with employees who arrive late, get tired, want time off, complain about conditions, get hurt, spoil batches and generally disrupt operations. The ability to move higher-quality product out the door faster would seem enough for any company to want to automate every process. In fact, trends suggest this may be happening. Management consultant Walter Bennis says that the factory of the future will have only two employees, a human and a dog—with the human there only to feed the dog and the dog there to bite the human if s/he touches anything (Paradies & Unger, 2000). Few would argue that automated control systems are not necessary in today’s complex industrial processes. However, is complete automation the best or most appropriate approach? While control system automation provides predictable, consistent performance, it lacks human judgment, adaptability and logic. Conversely, humans provide judgment, adaptability, experience and sound logic, yet are unpredictable, unreliable, inconsistent, subject to emotions and alternative motivations, and not biomechanically efficient. This raises two important questions: 1) To maximize system performance, should we automate humans out of the system? or 2) Do we maximize human input and lose efficient, consistent, error-free system performance? The answer is that the proper level of automation is likely somewhere between these two extremes and it is likely different for each system and situation (Haight & Kecojevic, 2005). This article reviews existing literature on automated control systems and human interface, and attempts to extend the work of Haight and Kecojevic (2005). The goal is to develop a method that helps design engineers determine how to minimize human error while maximizing system performance and better understand the right human/machine mix (Haight & Kecojevic, 2005).