Why a truly multipurpose robot is so important
Have you ever noticed that vanilla icecream basically serves as the basis for all other flavors and toppings. You make chocolate icecream by adding some syrup, you get rocky-road by throwing in some chunks. Sprinkles give it some color. A cone lets you carry it around. And yet all of these flavors and toppings are possible because of the vanilla base. But what if there were no vanilla? What if you just had the toppings and flavors each serving their own specific purpose, but in isolation.
That is what the robotics industry is like today. Right now robots are designed for very specific purposes. They only vacuum or they are for interaction or they just roll around. This is like chocolate syrup, sprinkles and a cone.
If you want a robot that can perform all of these jobs, a vanilla robot, they generally cost tens of thousands of dollars. The most affordable version at the moment is likely PAL Robotics Tiago at $50,000. But if you want a "perfect machine" there is the PR2 which is $250-$400,000.
At prices like these, the robots are like artisan vanilla. And that makes sense because each robot is basically hand-made to order because there is no volume. Without volume the price can't drop, but without some price drops there can be no volume.
But the vanilla robot is so necessary because it can serve as a foundation for development. With most robots being custom (flavors and toppings) there is little progress because each developer is working with their own special system, Because of this basic components and softwares have to be rebuilt for the specific machine as opposed to built upon. A vanilla robot serves as a platform that move the industry collectively ahead.
That is what Jerry was designed to be. He is the poor-man's PR2. The Model T of robotics. Simple and robust, but affordable. Jerry has the capabilities to be a vanilla robot, but we still need to build the software foundation for others to grow from.
It is official. Our Kickstarter has launched We are taking orders for the development versions of Jerry as well as preorders for the home version when it is completed. If you aren't ready for a robot yet help us create Jerry by checking out one of the other perks on our campaign
The Apple II is widely regarded as the first true personal computer. A computer that anyone could get a hold of and use. But what was it that made this device so popular that it created a new industry?
The first benefit was cost. At $5069, in today's dollars, it was a steal at a time when computers were behemoths costing tens or hundreds of thousands of dollars. But Wozniak was able to achieve that price-point, by simplifying the computer is ways that had never been realized before.
Wozniak never used two parts where one would do. He was not lazy in the design. He figured out how to push everything he could out of the circuits of the time. Sometimes he sacrificed expensive hardware and replaced it with affordable software. Functionally it was more difficult to achieve but it was worth the effort to keep the price down.
Wozniak also made the grand decision to use devices that people already had to interface with the computer. He used the TV as a monitor when it was standard to have a purchase a monitor. This was a simple solutions to an expensive problem.
Admittedly the Apple II was not really interesting to anyone expect hobbyists until spreadsheets were created. But that does not change the fact that the device was still more usable by a wider range of people than many other options. It was one of the first devices that made computers easier and more useful.
The Apple II was so successful because its design was simple and competent. There was nothing wasted. Just everything maximized. But was was the greatest part about it was that it was created by a 26 year old kid that did not have a degree. Wozniak flew in the face of conventional practices and created the first product in an entirely new industry, that others had tried and failed to enter.
Very often great products are not technologically revolutionary. They are not something akin to Einstein building a starship using his theory of relativity. Instead they are often the convergence of a set of well known industries and technologies in a way that was never before considered.
The internet was the convergence of computing and human communication. The personal computer was the convergence of transistor technology and interface design. The automobile was the convergence of metal processing, assembly techniques, and large supply of gasoline from kerosene oil. Robotics is right at the tip of a convergence.
The robotics industry has been stinted for decades by intelligence capabilities. Stupid machines were not getting smarter, so there was no way to introduce them in a significant way to consumers. But today AI and machine learning are progressing at an exponential rate and are receiving support from juggernauts like Google. And not only is AI being pursued more, it is being implemented.
Google's DeepMind already has astounding learning capabilities. And it is being deployed into more and more areas. They began with identifying photos on the internet and are now moving into analyzing medical records.
AR and VR
Another technology that robots have often been lacking in has been sensing. They have utilized lasers, ultrasonics, and any number of other technologies. But none has been well established or proven. But with the advent of AR and VR sensing of the environment and have a computer understand how to interact with it is developing very quickly. An entire industry is building whose primary job is to figure out how to reliably sense the world cheaply. Before today, that mission was relegated to college research labs for the most part.
This affordable sensing technology will be able to be implemented onto robots which need to sense their environment. And not only will the technology be affordable but it will be every improving because it is supported by several industries. Not just one robot startup.
Next, electric cars. Sensing and computing have moved far enough that it is possible to create a thinking machine that could perform some interesting tasks. The trouble is that the body of the machine is still highly expensive. The complexity of the mechanics of a robot are so great that the price for raw materials can be very great. Items like motors and batteries, which don't follow Moore's Law, like computers have a rock bottom price generally determined by volume. The robotics industry today does not have the volume to create any scale cost benefits. But there is another large electric machine that is gaining popularity.
Electric cars are going to create the demand needed to bring down the price of materials like copper and lithium for motors and batteries. The technologies that electric cars use for locomotion will also parallel those of robotics.
All of these industries are creating a component for a robot. Machine learning is supplying the brains. AR is going to create the sensors. And electric cars are going to provide the body as motors and batteries.
Any robotics company that exists today is going to have a jump on any competitors as long as they realize that they do not have to develop an entirely new technology to be successful. They simply have to combine the technologies that are progressing around them.
Not to sound self-serving, but this is the entire mantra at Slant. We are working to use the off the shelf technologies from all these other industries in our robot. To create them again independently will simply be a waste, because your new sensor tech would soon be eclipsed by the cameras in the new Hololens, or something similar.
Robots are a highly multidisciplinary technology. But now several industries are building on each particular component so that one company does not have to. Robots are the center of the technology convergence.