Engineering the Future of Work
Sometimes a key statement becomes a memory that last a lifetime.
I had been selected to become an engineering co-op student in my hometown at the local plant. I had spent the summer working on the shop floor before I headed out to college and the company had me rotate through almost every operation. It was a great place to cut my teeth. They had induction welding, tig welding, milling, drilling, turning, grinding, paint, and assembly all under one roof. When I returned from my first semester, I would be considered part of management and no longer be able to directly work on the shop.
It was on the send-off that I sat down with the Engineering Manager with whom I'd be assigned to when I returned. He said two (2) things that I've never forgotten.
Statement #1:
As an engineer, you know that you may be able to fix something to 100% or 100% of the cost. Your goal is to make it work 90% well at 20% of the cost.
This was a profound statement that epitomizes engineering. It is not always about perfection. It is balancing function vs. time vs. cost. Engineers commonly apply safety factors to ensure function is achieved while saving time and cost getting it perfect.
Statement #2:
If you can eliminate your job, you'll always have a job.
This statement has always carried weight with me. The job of an engineer is to improve, to automate, to eliminate waste. This has always come with a heavy weight as well. When you do your job well, it has a personal impact on yourself and people that you may know. It causes dislocations of people and friends. As I have told people about these statements over the years, I have had pushback stating that "How can you be a job killer?". To me this is looking at the jar half empty rather than the jar half full.
With the advent of the telephone, a major job growth contributor was the switchboard operator. The challenge was that as the lines scaled, the complexity increased and more operators were needed. This actually increased cost until this was automated as described in Milton Mueller paper. The automation forced thousands to be dislocated, however, modern communications would not exist without it.
McKinsey provided a very interesting article about the automation of work with up to 30% of work being automated in the upcoming years. It is a fascinating article with a good balance where some may see it as a doomsday picture and others as tremendous opportunity.
Working the engineering world, along with my key lessons as a young co-op, I see this as a tremendous opportunity. Today, automating engineering processes is not as hard as the past. It does not require extensive coding. In design, I estimate that 50%-80% of work could be automated today by capturing the engineering intent in templates.
The templates then provide a jumpstart to adaptations of products to new scenarios. We are not inventing new systems all of the time, rather we are improving upon them or adapting them to new design criteria. With a seemingly shrinking engineer talent pool, automation is a key strategy that companies should be employing.
In addition, engineering iteration today is primarily a manual process. Engineers provide the key inputs to the design. They provide the knowledge and intuition to look across the potential design space to define the next set of parameters to input.
With lower cost of compute power, the shrinking talent pool of engineers, and the ability to tie machine learning to simulation, the engineer should no longer be the sole input of the parameter sets.
The engineer should guide the parameter sets and constraints but allow the complexity of the design space be iterated upon by agents / robots that can define the strategy faster than humans. Yes, the machines will win here.
This is already taking place in IC design and wire harness design where the complexity has overtaken the speed at which humans can determine the best design.
Our world of engineering will dramatically transform over the course of the next decade. The future of work of engineer lies more and more in the machines.
My goal is to help automate the current generation of engineering work so that your next project is different than your previous one. I'm not worried, when I do, there will be something else more complex to work on. The problem set will change. It always does. I am excited to see what those next set of problems will be. I can guarantee that they will be more interesting than the current set.