The Semantics of Robotics
These days, few topics are discussed and debated as much as that of robotic automation. What jobs will robots replace? What industries will they benefit the most? Are they dangerous, or how will we make them safe? These are all relevant and important questions that deserve the many discussions already in progress, but I think a good place to start is to talk about some of the history of robotic automation, and how robots are changing and improving.
The first thing to note is that there are many different types of robots, with many different applications. In fact, robots span such a wide variety of shapes, sizes, and capabilities that few people agree on the definition of the word “robot.” Most specify robots as “machines that can be programmed by a computer, capable of automatically carrying out a complex series of actions.”
Others definitions hint at the notion of intelligence, or even a resemblance to a living creature, such as a human. Alternatively, I once heard a definition stating that any machine with three or more independent directions of motion (also known as degrees of freedom, or “DoF”) can be considered a robot! Because of this broad range of definitions and the many types robots currently being developed, for the purpose of this article I’m going to specifically focus on robots used for manufacturing applications ― the original and most common application of robotics.
Taking it From the Top with Traditional Industrial Robotics
If you look back through history for the first sign of automation being introduced into the world, you’ll land at Ford Motor Company in 1948. As the company was ramping up a new production line to compete with Chevrolet, Ford’s then Vice President Delmar S. Harder realized the need for improved material handling between production stages and famously proclaimed, “What we need is more automation!”. This statement is credited as the first instance of the word “automation” being used in the context of manufacturing.
After that, it was more than a decade before actual industrial robots were brought to market. In 1961 George Devol and Joseph Engelberger founded Unimation, Inc. and integrated the first industrial robot, the Unimate #001, into a General Motors assembly line in Trenton, New Jersey.
After a few more early advancements in computing and robotics, such as the invention of robust industrial computers known as programmable logic controllers (PLCs) in 1968 and the creation of the 6 DoF “Stanford Arm” in 1969, established players started to enter the space and build their own robots. And so emerged companies like ABB, Fanuc, Yasakawa Motoman, Kuka, Kawasaki and others which drove the industry forward and became responsible for creating what many of us now think of when you picture a vehicle production line in a modern factory.
This is the phase or type of automation that I like to refer to as “traditional industrial automation,” or “industrial robots”. Traditional industrial automation can be characterized by large, expensive, deeply-integrated robots that you’ll find in a modern day car factory. These robots have become incredibly reliable and high-performance, capable of lifting an immense amount of weight and moving at terrifying speeds. However, these robots are still “dumb”, controlled by now archaic PLCs rather than modern software. But this is not where robots are heading today. Where we’re going, robots are capable, smart, and accessible.
The Dawn of Modern Advanced Robotics
The advancement of robotics directly correlates to the emergence of cheap and powerful computers known as microcontrollers, which gave the ability to run sophisticated software on-board a robot. One of the first, if not the original “modern” or “advanced” robotics companies to form was iRobot, which was founded in 1990.
Twelve years later in 2002 they released the now famous Roomba robot vacuum cleaner. Though drastically different from industrial robots, the Roomba was one of the earliest mass-market robots with integrated “intelligence” to navigate around your floors picking up dirt.
Shortly after that in 2003, Kiva Systems was founded with the goal of building autonomous mobile robots (AMRs) to drive around warehouses helping to move inventory from A to B. Sound familiar? They were later purchased by Amazon in 2012 and have since become Amazon Robotics.
As you move forward through the 2000s the industry continued to evolve and branch out, with more and more companies entering the arena ― Willow Garage in 2006, responsible for the creation of the increasingly popular open source robot software framework, ROS, first released in 2010. Or Universal Robots in 2005 and Rethink Robotics in 2008, both early creators of collaborative robotic arms, or “cobots”.
However, while there was a lot of movement in the robotics industry during the 2000s, much of what was brought to market took years to develop, or wasn’t yet ready to be deployed in the real world. And so during this period, a lot of the innovation in this space didn’t happen in industry, but in research labs where the unique combination of accessible low-cost robots and open source software paved the way for research-level development.
Since then, these technologies have finally become capable and cost-effective enough to reach the global market, and are seeing their way into real-world environments. Enter the age of advanced robotics.
What Lies Ahead
Although it might not be clear when the shift happened from traditional industrial robotics to modern advanced robotics, it’s obvious that both our technology and attitude are ready. This new generation of robots will fundamentally change how the world operates.
Industrial robots ushered in the era of factory automation where there was a clear dividing line between robot and human work. New robots will be semi-intelligent, and will be both safer and easier for humans to interact with. They will work side-by-side with us and can be trained on-the-fly without any higher-education or knowledge. Over time, these robots will almost certainly become ubiquitous in factories, and in most other settings.
We’re about to enter a phase of acceleration, where new technologies such as deep learning and advanced manufacturing, and peripheral industries such as self-driving cars, will explode the number of ways and places we can use robots. Buckle up, the ride is about to begin.