Category : Artificial Intelligence
The factories of today can be compared to futuristic hives of automation where industrial robots imitate the movements of workers and possibly also its intention
Robotics and artificial intelligence nowdays operate not only faster and more reliable than people, but they also perform perfect intelligence tasks that exceed the skills, like for example the microscopic Assembly precision. It is also true that many of them are not as smart as they seem, or rather, still, more skilled people, are scheduled to run a limited number of tasks but the importance of artificial intelligence is immersive. Many robots pose a hazard when operating close to the people and must be located in isolated areas or operate under Regulation to protect operators.
As technology advances and costs decrease, and as manufacturers discover applications for which the algorithms of artificial intelligence today (AI) can solve complex decisions, AI is increasingly a specific place in manufacturing. As it becomes ubiquitous, the future of AI in manufacturing is already feasible in emerging markets, put in relief the current improvement in sensory capabilities and its use outside the production line and predicts what is going to need and when.
1 One a more widespread use of robotics artificial intelligence in new markets
Manufacture requires a large amount of investment and frequently provides low profitability. For this reason, has moved much of the manufacturing to countries of low wages where labor costs are so low that invest in AI and automation was formerly difficult to justify. Currently, as a result of the increase in salaries and living standards in countries such as India, you can easily insert the robotics and artificial intelligence. China is already devoting a considerable investment is computing artificial intelligence for manufacturing and e-commerce. China has seen the future of robotics and artificial intelligence and all the benefits and risks of artificial intelligence, the questions is when other highly developed countries will follow the same trand.
As we have seen how workers in industrialized countries were complaining of the loss of jobs as a result of automation, we now see the same phenomenon in the China factories. It is true that in the short term, many workers will be replaced by robotics and artificial intelligence, but is intended to hold them to make a more sophisticated design, programming or maintenance tasks. The ultimate goal will be to develop applications of AI that not only have intended the automation of tasks, but they also carried out the process of the business in its entirety, as to make products by the specific requirements of a client.
2 Two best sensors on the machines to achieve greater safety in the workplace
Artificial intelligence coding dates back to the 1950s but has not been accepted widely to the development of machine learning algorithms applied to data sets to discover meaningful patterns, without having to do this one Express programming. Michael Mendelson, one of those responsible for the development of the curriculum of the Institute NVIDIA Deep Learning, says: “without flexible algorithms, computing artificial intelligence can only do what we asked them to do. Many tasks, especially those related to perception, can not become instructions based on rules. In the manufacturing sector, some of the most interesting applications shortly will include the perception”. This could mean that robots are more capable and resulted ideal to interact with and receive instructions from the people.
The vision is one of these applications. Designing cameras with a sensitivity higher than the human eye have been a fairly easy task. Artificial intelligence now offers us the possibility to give increasingly more meaning to the images. For example there is a startuper Andrew Ng, a veteran of Silicon Valley, occupied in resolving problems of manufacture such as high precision quality analysis. He has developed computer vision tools that can find microscopic defects in products such as printed circuit boards of a resolution that is undetectable by the human eye using an algorithm of machine learning has been tested using a very small number of sample images.
This is a challenge to the micro-level. A macro-level problem is to form a robot for you to detect what is happening in your environment and that can thus avoid disturbances or hazards. It is a situation analogous to the vehicle led auto whose use in the world is getting closer. The factories they may be equipped with a forklift and smart conveyor driven auto that move materials and finished products.
The robots tend to be installed in a fixed location but at risk of colliding with other objects or people that led by the place in which they operate. Machine vision or motion sensors can get to halt its activity in case of possible danger. There is also more and more interest in the collaborative robots, also called cobots, that can operate at working with people. There is artificial intelligence coding that makes possibility to receive instructions from humans, new instructions that were not foreseen in the original programming. So robots and people should make use of a common language, something that resembles more and more normal speech. This is a concept already it has shown at the University of Rochester (USA) and the Massachusetts Institute of Technology (MIT).
“The human species has accumulated experience of thousands of years exercising verbal communication while the written form is much more recent and sometimes more farfetched.”, says Mendelson. “Talk to robots allows us to communicate concepts that, through texts, perhaps could not be expressed so clear.”
3 The robotics and artificial intelligence in the chain of logistics and other possibilities
Artificial intelligence coding makes robots more capable and facilitates our collaboration with them. Its impact also may be in sectors that are not related to robotics. In the logistics chain, for example, algorithms can detect patterns of demand for certain products in precise moments, in different geographic markets and specific socio-economic segments and take into account at the same time cycles macroeconomic, political situation and even weather trends. You can get to improve forecasts of the market demand and his subsequent supply of raw materials, staffing levels, better financial decisions, stock, maintenance of equipment, and consistent energy consumption.
Importance of artificial intelligence is also useful for the predictive maintenance of the equipment, sensors that verify the operating conditions and the performance of equipment of factories, as well as to predict faults and malfunctions. These sensors can also recommend or take preventive measures. SOM Shahapurkar is director of machine learning in FICO, with over 40 years experience in the commercialization of AI and says that in other sectors, this is a common practice. It adds that its use has spread to different areas, both to the generation of complex emails directed to consumers or the prediction of failure of blades in towers of Facebook and Google servers.
Much of the data will come from sensors in the same processing equipment and not only at the factory. Sensors may also be located in the centers of the suppliers, dealing with check inventory of spare parts and other inputs front and deal with of the monitoring of the quality problems of products at distribution centers or in retailers.
Artificial intelligence today can give guidelines to help manufacturers to predict demand before they manufacture products in preparation. In 2010 the Professor of computer science Johan Bollen along with several colleagues from Indiana University demonstrated that algorithms could read and interpret the climate of opinion on Twitter with such precision that they allowed predicting the movements of stock markets. This widespread sense of analysis could be used to predict the demand for certain products or even specific brands, says Bollen, “especially today when consumers frequently communicate their impressions to chats. So Google and Amazon AI assistants should be used”. Much of the recent work developed by Bollen focuses on the influence of social networks in the climate of political opinion, although he has also studied the behavior of consumers.
Even so, advocates of artificial intelligence say that this technology is just an evolutionary form of automation, an inevitable result of the fourth Industrial Revolution. Artificial intelligence can be an effective tool in manufacturing, to achieve better development and produce things more cheaply. Even so, the truth is that human ingenuity cannot be replaced to cope with unexpected changes in demand and tastes. Neither nor to decide if something is going to make or not.