Robotics engineers are proud of their invention. They are talking about neural networks, expert systems, Artificial Intelligence and how all of these technology will revolutionize the society. What didn't know is under which cases robotics automation can't be used in reality. The surprising fact is, that apart from the progress in the last decade, there is no single robot on the market which can be recommneded for household or industrial application. This is surprising, because lots of new software was written, modern computer hardware is available and according to the latest youtube video, robots are able to walk on two legs. The answer why robotics has become powerful and useless at the same is given by a short history lesson.
In the late 1970s, lots of electronic devices were invented. Electronics was the precursor to computer technology and it was used to power the first transistor radios, supermarket checkout machines and automated elevators. The introduction of electronics was a revolution which helped all the companies and private households to save lots of time. The simple but effective barcode system used in the supermarket as an identifier for products is a good example how powerful electronics is. The system increased the productivity a lot.
The interesting fact is, that since the electronics revolution in the 1970 no further breakthrough was there. The logical next step after transistor based barcode readers would be a robotic checkout. This kind of innovation was never made. Many failed robotics project were tested out, but the engineers were not able to repeat the success of the electronics boom in the 1970s. The robotics revolution from the 1980s to the year 2020 is missing. It was an innovation which never took place.
If today a household or a companies plans to use the latest time saving technology available on the market, they will decide for the well working electronics devices released in the mid 1970s. Such technology has a clear advantage over other alternatives and in contrast to robotics it can be used in reality with an added value.
The reason why robotics never replaced normal electronics in mass production is not researched very well. What we can observe from the literature is, that some attempts were made and they always failed. The first one was the Unimation company founded in the 1960s, in the 1980s the Helpmate robot was introduced, and in the 2010 the Rethink company developed the Baxter robot. Additionally, the Volkswagen company tested in the mid 1980s in the Halle 54 project the newly developed Computer aided manufactoring which includes robotics and the latest innovation from the year 2019 is a self-driving car which is tested on the roads to solve the traffic jam problem. All of these attemps have in common that they have failed. That means, they are producing lots of costs, but didn't improve the productivity. On the long run, they are canceled, and if not they will do in the future.
Like i mentioned before the reason why isn't researched very much. From an outside perspective, the Helpmate robot which was developed in the 1980s was a great hospital transport vehicle. It was equipped with sensors, had an onboard computer and advanced software was written. But something was missing. In contrast to electronics innovation, the product was never loved by the customers. It was perceived as something which doesn't improved the situation but make it worse. The same is true for household robots which are looking fine on the first look, but in reality they are struggling.
It seems, that the jump from normal electronics devices which are state-of-the art technology to a modern robotic based technology is much harder than expected first. It seems, that simple electronics devices are more powerful they look on the first attempt and robotics is less powerful than it seems. A look into the reality have shown, that it make sense to replace a robot with it's electronic counterpart from the 1970s and this will increase the productivity drastically. That means, the latest innnovation available wasn't invented in the year 2010, but it was produced in the mid 1970s before the advent of the microprocessor.
I'm not the first one who is describing the productivity paradox. It's a well known fact that until the 1970s the productivity in the economy has grown constant but since the 1980s it was no longer growing but became even lower. According to the history of technology, the 1980s was the decade of the microcomputer. It was the first time, that microprocessor based devices like the Commodore 64, the IBM PC and software products were introduced in the mainstream market. Even if the technology is great by itself, it struggles to increase the productivity in the economy. In most cases, computers and robotics is useless for companies. What the are prefer instead is normal electronics device from the mid 1970s.
From a technical point of view, computers are superior to the 1970s electornics. A single computer can be programmed for many purpose. On the other hand, the advantage can't be used for increasing the productivity in automation. This kind of paradox remains an unsolved problem. Today's computers are much more powerful than the technology in the 1970s and advanced software was written, but all these technology is not able to automate simple tasks like a checkout terminal in the supermarket.
Perhaps it make sense to mention the limits of 1970s electronics. The basic idea behind simple barcode scanners, plain calculators and early sensor systems was, that the technology is limited. A barcode scanner can do nothing apart from reading optical information. To use the device in an industrial context, a human worker is needed which was trained to use the device. Nearly all electronics devices have an obvious limitation. In contrast, the computer technology is promising endless capabilities. The idea behind a personal computer is, that the device can be programmed in any direction. And in theory, the computer can be programmed so, that the robot will do the work instead of the human worker. This kind of promise was never fulfilled.
To understand the productivity paradox in detail we have to investigate who well productivity was grown in the past. Until the mid 1970s the engineers were successful in developing new technologies. They introduced better cars, better automation techniques, and they invented new household machines for example an automatic washing machines. All these devices have in common, that it was easy to convince the customer to buy it, because the advantage was obvious. Using a washing machine is more productive, than not using a washing machine.
And with the advent of the first computers in the 1980s the development stopped. The engineers have developed advanced computer chips, but they struggle in creating useful products which are superior than machines from the past. Today's cars, today's households and today's factories look the same like in the 1970s. The latest time saving technology is in charge, but what is missing are robots which can increase the productivity more.
A naive assumption is, that the technology is available but the people don't understand it. The opposite is the case. Many companies have experimented since the 1980s with robotics and factory automation because they want to profit from the latest advances. All of these robot projects have failed. They were tested on a small scale, and no sign of advantage was visible. It seems, that the computer can do many things, apart from increasing the productivity in the economy.