The history of automation is a very interesting topic, being a history that has spanned thousands of years. However, in that amount of time, there have been very few people who have studied it. This is very unfortunate, because, without understanding our past, it’s hard to look into the future. I’d now like to take a minute to look at the known historical facts of automation with you and try to put them together into a system.
To focus specifically on automation, let’s remember the definition: “The use of technology (machines, control systems, and information technologies) to enhance or completely replace manual human operations.”
That is very a broad definition with many things that may fall into it. Even in a primitive society a human used sticks, stones, and bones as tools to increase the efficiency of his work. He also domesticated animals like dogs, cats, and horses, which expanded his abilities even more. But all those objects cannot quite be considered as automation, because they are not created technologies. Those things were borrowed from nature and used with minimal changes. Technology, on the other hand, assumes something non-trivial, which is created by humans themselves. It can be defined as a product of human work, which can then be used to create other products.
The first automations used by humans were technologies such as the bow, wheel, mill, and blast furnace. These are objects that appeared thousands years ago, and we could call this first period of technology the 1st Era of Automation. Many of those automations were so primitive that several people would not agree to call them as such. However, let’s look at little closer...
First of all, automation can be done in two different ways--by usage or delegation. In the case of usage automation a function cannot produce a solution by itself. A human operator must be involved throughout the entire process from the beginning to the end. Such a function is more correct to be called a “tool” or “passive resource” (or sometimes just a “resource”). In the case of delegation automation, a human is able to set a specific function to carry out a task, leave, and then come only at the end to obtain the produced results. The automation in this case must be able to perform a sequence of operations by itself. Those functions we can call “active elements” or “active resources.”
Looking at these two types of automation, we find that the early primitive tools of the 1st Era of Automation were passive resources and can be claimed as automations. The bow, wheel, mill, and blast furnace were tools used to bring about solutions for their operators, and thus, for this reason, we can say that the 1st Era of Automation began with these inventions.
Initially, technologies of the 1st Era of Automation were used only for the Act functions of the OODA loop and included actions such as “lift”, “move”, “put down”, “kill”, and “cut.” Later, with the progression of human knowledge, we then began to see the automation of Observe functions. This included passive tools such as magnifying glasses to look at fine details, sand and sun clocks to measure time, and flags to determine the speed and direction of wind. Centuries then passed before humans approached the automation of Orient functions. That happened, perhaps, with only with the invention of the mechanical calculator by Blaise Pascal in 1642. Finally, the automation of the Decide function has only come in the last hundred years with the invention of computers during World War II.
By looking at this evolution, the 1st Era of Automation, which I call “primitive automation,” lasted for thousands of years and ended only in the 40s of the previous 20th century. That era is characterised by the automation of individual management functions performed by individual people.
It was only with the the invention of the computer that we began to see a change.
The invention of computer technologies allowed humans to master all functions of the OODA loop and start the implementation of the first autonomous, intellectual active elements. However, management systems still relied mostly on manual work and automation was used only in isolated islands. The majority of functions remained non-automated and were performed manually by human workers. This marks the beginning of the 2nd Era of Automation, and it includes fragmentary, partial, or narrow automation. That era ended not so long ago--just at the end of the last century. However, that’s completely true only for developed western countries, where computer technologies sneak into almost every area of human life. There are many other places in the world where that is still not the case.
Today, developed countries have transitioned into the 3rd Era of Automation, which includes complete or wide automation. In most organizations it is now very hard to find a person who does not use automation of some sort somewhere. Even workers such as janitors, porters, or plumbers more often than not use computers and other automated tools in their work, at least to keep track of their orders and communicate with customers. In this 3rd Era of Automation, with automation at such a high level, almost all active elements in organizations have become more or less covered by automation. Most interactions have started going through computer systems. The total number of workers in organizations decreased and their productivity significantly went up. Finally, some professions disappeared completely, as they were replaced by technologies. With so much automation, there are only a few functions today that could not be automated as they heavily depend on advanced cognitive or creative human abilities.
Today the main barriers for further automation are caused by a few key factors:
- Technical complexity, or total inability to automate at the current level of technologies. Typically, this problem exists in areas with high uncertainty/variability and the absence of clear standard solution patterns--areas such as research, the development of new technologies, business management, and work in uncontrollable, chaotic environments such as a modern metropolis.
- Economical rationale. Some jobs, such as cleaning houses or cooking and selling hamburgers are relatively cheap for businesses because of low-paid workers and it could be quite expensive to introduce replacement technologies.
- Human interactions. Modern systems still experience troubles communicating with humans using human languages (and not all the people are comfortable to speak with computers through a technical language). This causes mistrust toward computers. Because of that, in the areas of human interactions, such as HR management and customer service, human workers are still widely used as mediators between other humans and computerized business systems.
However, we can see how those barriers are quickly disappearing right in front of us. Advances in robotics, Artificial Intelligence, and the massive introduction of complex sensing technologies have increased automation capabilities exponentially. Cost of automation systems is rapidly going down. Human-Machine Interfaces have become more sophisticated, and people are more computer savvy today than yesterday. The level of trust toward computers is constantly increasing.
Today the human society stays at the doorstep of the 4th and, possibly, last automation era: the era of comprehensive or total automation. In this coming era, automation technologies will replace humans in most areas of production and services, and only a few areas may still remain under human control. However, as natural barriers for further automation disappear, other artificial barriers could be introduced using political means. People would still need jobs. What can be done to employ jobless workers? How can they support themselves? Those may become very tough questions. But this is not in the area of my expertise, so I’m not able to give any meaningful comments.
On the other hand, I’m very lucky (from an automation specialist point of view, of course) to work in the industry, which very soon won’t be able to sustain itself without comprehensive automation. And because of that, it doesn’t and won’t have any strong political barriers. I’m talking here about the Mining industry. Natural resources in easily accessible locations on Earth are being depleted, but the demand keeps growing. In order to keep mining, mining corporations must start operations in deep mines and underwater on the ocean floor in the next two to three decades. After that, in the not so distant future, mining shall be done on other planets and asteroids in space. Creating comfortable and safe working environments for humans there will be extremely complex and expensive. And, believe me, there won’t be too many volunteers to spend their lives in places like that. Thus, there won’t be any viable alternatives to comprehensive automation in mining. We are talking here not just about simple operations. Imagine mining on a remote asteroid at the edge of solar system. To operate, automated mining machines shall do construction, mining operations, processing, transportation, and equipment maintenance. That will also require long-term planning and crisis management tasks. In order to do this, we are looking to automate all horizontal areas as well as all vertical management levels up to the strategic level. Searching for solutions to solve that problem became a starting point for my research on goal-oriented management and comprehensive automation.
To compare the different eras of automation I composed the table below:
Characteristics to compare
Approximate end time:
40s of 20th century
80-90s of 20th century
20-30s of 21st century??
Key differentiation factors:
Manual work with automation of individual management functions
Automation islands, domination of manual work
Domination of automation, islands of manual work
Complete replacement of routine manual work by automation
Machine-tools, simple sensors, arithmometers
Automated conveyors, SCADA, Management information systems, remote control systems
Integrated enterprise management systems, autonomous robots
Completely automated production and service organizations
Paper workflow, office clerks, controllers
Customer service, strategic planning, low-paid jobs
Art, research & development organizations
We are about to transition into the comprehensive automation era. Social needs and the critical mass of technologies are already getting close to that level. I just worry about the lack of fundamental understanding of how all of it shall work. There is still no theory to support everything.
During my short lifetime I witnessed the transition from the partial to the complete automation era. And, I have to say, that the chaos in the heads of automation specialists caused a lot of negative results. I’m afraid that the chaos will remain. Only God knows what could happen when humans decide to completely trust their lives to machines. I think it makes sense to figure it out before jumping there.
That shall be all for today. In my next posts I’m going to discuss in detail a few aspects of comprehensive automation, starting from the new modeling methodologies.