In order to comprehend the rise of robots, it is important to understand why they are not already a huge part of our lives. Researchers have consistently predicted a robot revolution for the last fifty years, but why has it not yet happened?
To understand this, we need to first look at the brain of a robot, or its processing power. As you will see from the meteoric rise of processing power and the reduction in cost in the graph on the next page, we truly have been following the trajectory set out by Moore in 1965.6
We have developed the computing power to control the movement of robots over the last fifty years. Such developments have so far been implemented in large repetitive manufacturing plants that require a robot to perform the same repeated task. This is especially useful in car manufacturing. In reality, the processing power required to make standard repetitive movements is minuscule, as today’s digital wrist watches could complete such tasks without a problem.
Thus, if we have the processing power to move a robot in any which way we desire, what holds us back from creating a robot
that completes every task we want it to? One issue is mobility, which will later be discussed, but the most important issue is the authentic dynamic environment to which robots still find it very difficult to adapt and interact with. While a robot can perform repetitive tasks, or NASA likes to put it, “Robots tend to perform as well or better than humans when the tasks and conditions can be reliably predicted”, it is woefully underpowered when it encounters authentic changing environments and tasks.7 It is only recently that processing power has been available for us to effectively replicate our basic senses. In order for robots to make a profound impact in our world, they must be able to understand, process, and react to the variations of the surrounding environment.
Although robots face this difficulty, there are an emerging number of technologies that that are turning robotic environment processing into a reality. Many of these technologies have grown out of the mobile phone revolution, which has pushed a substantial increase in high quality, low cost microelectronics over the last ten years. The first of such microelectronics was that of low cost, high-resolution cameras that allowed low cost three-dimensional stereo vision. The first robots with three-dimensional stereo vision began to perceive depth, an extremely important ability for performing complex tasks in a dynamic environment. Imagine trying to navigate through a crowded space using only a two dimensional view and the challenges that the camera’s limited vision provides.
When Google began to develop the self-driving car, the company needed a technology with more accurate vision to gather precise images. Google opted to use LIDAR, a fancy acronym representing a laser that finds objects in the range of a 360 degree arc. The full form of it is “Light Detection and Ranging” or “Laser Imaging Detection and Ranging”. Think about LIDAR as an extremely fast spinning lighthouse that illuminates everything in its path, and carries the cost of several thousand digital cameras.
Needless to say, a good deal of these technologies, as well as the technologies to intelligently interpret them, will combine with the currently available processing power to make the “thinking and sensing” side of robotics relatively simple. This leads us onto the two real problems with current robotics technology, power and movement. The power problem is exactly the same as that of a mobile phone. As processing power of a mobile phone over time has increased, battery technology has unsuccessfully tried to keep up. Presently, in order to power a decently large robot for consistent useful work, it must be either tethered to the mains power or powered by a mechanical engine that generates power. There are huge advances in technology in this area, the first being improved battery technology. Even though batteries unfortunately do not follow Moore’s law, they do increase their performance by around 5% a year. Secondly, we are making computers more efficient with the available energy.8 The other developing area is that of fuel cells that are just beginning to make their way into consumer products. Fuel cells will transform natural gas into water in one compact unit with the benefit of generating significant electrical energy from an energy dense fuel.
Evidently, a large number of the technologies that will be required for the robotic revolution are already nearly in place. So, what is missing? Well, it is really quite simple. We have yet to invent a better, more efficient, more compact, cheaper means of generating movement than the humble electric motor invented in 1821.9 This simple device is responsible for servo, a device that can accurately move in a circular 360 degree arc for driving gears or wheels. Motors are relatively heavy, large and inaccurate for the fine movements that robotics require.
It is not as if we have not tried to come up with a better solution for generating robotic movement. All sorts of actuators, motors that turn energy into motion, have been tested. Some have been quite successful in their narrow field of use. Hydraulic actuators have been used in industrial robots that require great strength, while air actuators have been implemented in more delicate settings. However, one will encounter each actuator’s strengths and significant weaknesses when trying to build ubiquitous robots.
Thus, in order to try and solve the limitations of actuators, some very clever scientists are working with advanced materials. We now have artificial muscles that can contract in the presence of electro-active polymers, basically electricity, or heat. Even though some of these new technologies can deliver relatively strong responses at low energy levels, the amount they actually contract is not enough to perform useful work. I suspect that once these new materials mature, they will indubitably drive the robotic revolution forward. Humanity is a good ten years away from this new technology’s incorporation into products, but when it is entirely incorporated by then, the face of robotics will be forever changed. It is fascinating that a number of these technologies have been inspired by biology. We must not forget that the human body is an incredible device. While we may lament that we do not have the robots that we were promised for hundreds of years, we are well on our way to replicating that which biology has been perfecting for millions of years.