We are not on the verge of a world run by humanoid robots with superhuman artificial intelligence. Rather, we have reached a point at which a combination of digital and physical assets can be effectively applied to an increasing number of industrial processes in a way that improves speed, quality, convenience and efficiency.

Recent breakthroughs in deep learning, a computationally heavy form of artificial intelligence, are particularly important from a technological perspective. For example, robots are now learning to complete industrial tasks independently through trial and error.

In a recent case, it took a robot eight hours to learn how to correctly pick items out of a bin with high accuracy. Manually coding this process to a similar level of accuracy would have otherwise taken about ten times as long.

Additionally, automation systems are increasingly connected to the cloud. This enables distributed learning where processes in one factory can be synchronized across facilities globally in a matter of minutes. This will result in significant savings in programming costs.

In the U.S., we are seeing rapid changes in the distribution and fulfillment of goods ordered over the internet. Customers are demanding faster order processing and delivery.

Major retailers are being forced to invest heavily in equipment to automate order fulfillment. This includes cameras that rapidly inspect and identify goods and packages, warehouse conveyance systems driven by order management software and robots that can move, sort and package items.

Warehouse automation systems will work 24/7 without breaks, functioning in “lights out” warehouses with almost no human intervention.

In the current economic climate in which growth is scarce and investment dollars are precious, most companies are only willing to spend on projects that promise a clear and relatively quick return on their investment.

While technological advances make new automation applications possible, adoption is also being driven by powerful demographic factors. In 2010, the European Union’s working age population (15-64 year olds) peaked and it is expected to be in decline indefinitely going forward.

Interestingly, two of the four largest industrial robotics manufacturers are European. The other two are Japanese and, not coincidently, Japan’s working age population reached its peak even earlier, in 1995.

It is not just Europe and Japan that are aging, similar demographic pressures are also being felt in the world’s two largest economies, China and the United States. In fact, the ratio of those out of the work force versus those in the work force, known as the dependence ratio, is on the rise for the entire developed world, as well as China.

Over the last several years, there has been much discussion concerning the potential “re-shoring” of industrial production back to the United States. If this is to happen on a large scale, automation technology will be the key enabler.

The cost of automation equipment is the same regardless of geographic location. As these technologies continue to advance, the importance of the wage differential between workers in the United States and those in emerging market economies, which have taken over much of global manufacturing in the past two decades, will be minimized.

Thus, the calculus behind investment decisions for industrial companies will change in a world in which production costs become increasingly equalized. The most important geographic factor may prove to be the location of customers rather than the local cost of production. This bodes well for the future of manufacturing in the United States.

---

Spencer Smith is Director of Research and Managing Director at Chevy Chase Trust, a wealth management and investment firm in Bethesda Maryland. Prior to joining Chevy Chase Trust, Spencer was Managing Director, Portfolio Manager and co-head of the Washington D.C. office of Fiduciary Trust Company International.

READ LESS