You hear this a lot in the news, automation is taking all the jobs away from the hard working manufacturing employees. Soon there will be nothing left for the working class to do. The articles usually spell disaster for the economy and society in general. I have always wondered who is writing these articles. I am curious if they have ever visited a manufacturing facility, let alone worked in one, because this is not the scenario I see playing out.
First, let me explain a little bit about myself and what I do. I have a mechanical engineering degree and was hired by a manufacturing company right out of school. We manufacture a large variety of building products, and the company generates 2 billion dollars in revenue yearly.
I worked for three years in one of our manufacturing plants as a process engineer. I have recently transitioned into a role designing systems in the new plants we are building. Both of these roles have helped shaped my thoughts on the man automation dynamic.
There are three main reasons we are very far from designing human free manufacturing plants.
Diminishing returns are phenomena in which the amount of increased output reaches a point where it rapidly slows for every additional input. Eventually, the time and money investment required is greater than the time and money gained from that investment.
Think of this as an exponential equation, with the amount of work required to completely eliminate humans from a manufacturing process increasing at a much more rapid rate as you get closer and closer to that goal. At this point in time assisting a human with automation is easy, eliminating a human is increasingly harder.
When I was working as a process improvement engineer one of my main focuses was to remove responsibility from the operators. Some of the things I implemented made it easier for the operators to run our equipment properly, but I never came close to actually replacing their role entirely.
Now that I am on our new plant design team I get to see this effort at our businesses highest level. We are now at the point where it has become extremely difficult to reduce employees. We are a highly automated process already with tons of conveyors, stacking equipment, and more photo eyes than you could count. The plants we are building right now have new technology integrated within them that replaces two operators per shift. This system costs roughly 8 million dollars! We had to spend 8 million dollars to remove 2 men out of a 20 man operations crew.
The really interesting thing is that this investment is cost neutral. This means theoretically we will not be saving any money doing it this way in the short or long term. The true value of this business decision lies in things like safety and other hard to quantify advantages.
This shows you how large of an investment it can take to automate, and this was the lowest hanging fruit. We are at a point where it will take so much time, effort, and money to replace the next operator there will be no advantage to doing so.
Cost of Failure
The cost of equipment failure is the second major factor preventing total manufacturing automation. Murphy’s law is ever present in a manufacturing plant, and in life for that matter. Murphy’s law simply states “What can go wrong, will go wrong”.
In a manufacturing plant, people are a valuable insurance policy. Depending on what role you play in in the plant your job may solely be to watch the equipment run and take action when something appears to be going wrong. The operator sits and watches the computer screens and machines and monitors all the variables. Temperature, speed, position, dimensions, and quality specs are a few of the things that must be watched.
The operator stops the product from being outside of specs and identifies when the machine is in danger of shutting down. If that person was not there the business could lose thousands of dollars in a matter of minutes. Putting that responsibility in the hands of automation can be scary and dangerous.
It may seem like it would be easy to place this responsibility in automation’s hands. But let’s think about what trying to drop one of these guys would entail. You have to create feedback loops that take these inputs and make changes on their own. Yes putting in sensors to do tasks seems simple but trust me, all of these sensors fail eventually. Failed sensing equipment means bad product or broken machines. To prevent that you have to put in more equipment to monitor the sensing equipment. Things have a tendency of getting very complicated very fast, and complexity is almost always bad.
It takes a huge amount of effort from engineers and other team members to fully automate an entire part of the process.
This is where Murphy’s law comes in, if a programmer or engineer forgets one small scenario that would cause catastrophic failure, no matter how unlikely given enough time it will happen. Equipment failures and bad product get expensive fast. Sensing equipment is valuable, but it works best when assisting people not replacing them. The operator makes sure all the equipment and automation is working properly. Operators quickly troubleshoot along the way and communicate with maintenance. Operators catch and document the scenarios that are unexpected so the team can prevent them in the future.
Long after all of the floor operators are gone, there will still be a thriving maintenance workforce. Every system no matter how simple or complicated requires routine maintenance. Bearings wear out, materials corrode and fatigue, lubricant and hydraulics need changing, and sensing equipment gets dirty. Maintenance is an art and requires far more thought and talent than most realize. In fact, I almost did not include this in the article because it is hard to convey the magnitude of maintenance. If you work in a plant as an engineer or manager you will quickly learn a good relationship with the maintenance team is critical to your success. Their willingness to help you can literally make or break your career. At this point maintenance of equipment requires a level of troubleshooting, dexterity, maneuverability, and knowledge that robots are a long way from and may never possess.
At this point, we are struggling to find financial payback from any more automation. This coupled with the potential costs of the failures no one anticipated provides job security for plant operators. Maintaining that equipment provides another set of challenges. Robotics and automation have brought in amazing abilities to manufacturing. Companies are able to produce better quality products at faster rates and longer periods of time due to advancements in these fields. Automation is welcomed in our plant because it allows operators to do their jobs effectively, but it is not going to eliminate them.