Technological shortcomings include the following:
– power supply (either have low speed and frequency of data, or need to arrange power supply);
– dimensions (not all sensors can be miniature);
– equipment calibration (reliability of readings);
– dependence on the data network;
– lack of common protocols and standards for transmitted data, which may make it difficult to process, integrate, and analyze data even on a single production scale (in February 2022, the new ISO/IEC 30162:2022 standard was released, but the transition to uniform rules will still be difficult);
– vulnerability to external attacks and subsequent data leaks or intruders gaining access to hardware management.
5G
You’ve probably heard of the 5G. That this is a breakthrough in communications, and no new flagship can be a breakthrough without the 5G. After all, without it it is impossible to look at the smartphone new series at 4 or 8K. Therefore, you need to buy smartphones only with this module and pay 100—150$ more than the version with the 4G module.
However, very few know that the standard itself was not designed for video in YouTube or TikTok, but for the scale development and implementation of digital services. Its «chip» is flexible combination of ultralow latency (URLLC), high speed (eMBB) and reliability of the communication channel (mMTC), depending on what is needed by a particular subscriber.
Basically, it’s a connection for the IoT. It may not be entirely suitable for an industrial IoT, but for a smart city, healthcare, and industrial enterprises in the city it is the ideal option.
So, what is the difference between 5G and 4G/LTE?
– Eight times better energy efficiency
– 10—100 times the speed
– 100 times more subscribers per base station
All those who are engaged in digitalization in production, and even just implementing ACS TP, know that the main problem is to organize data transmission to or from sensors. The solution of this issue in accordance with all the rules of the company is sometimes several times more expensive than «iron» and software.
Additionally, I hope that with the development of 5G technology, this problem will become less and less relevant.
In addition, the development of this technology will also help the implementation of more advanced IT systems, especially MES, APS, EAM, BIM. More about them – in the next chapter. All these systems need information from sensors without human intervention.
However, there is an unpleasant moment for many. All this will require other competencies from the employees. This means that the «optimization» of the organizational structure and the increase of social tension will begin.
6G
China and the US are already developing standards for 6th generation networks. However, why?
To ensure further growth of smart device deployment! 5G still has limited capacity.
Some sources suggest peak speeds of up to 1 Tbit/s. Average speed of several hundred Mbit/s. The average signal latency is 1 ms, which is useful for applications that require minimal latency, such as autopilots and virtual reality. The number of active devices that can connect to 6G per unit time will also be several times higher than 5G.
«The 6G Era will offer new possibilities for creating brain-computer interfaces», says Dr Mahyar Shirvanimogaddam of Sydney University. An example of such development is the electronic chip for paralyzed and people with CNS disorders, which is created by Elon Musk’s startup.
In this case, the 6G has one great advantage – it is possible to upgrade the existing 5G towers for its implementation, while the 5G had to build new base stations.
It is now believed that 6G may be introduced in the early 2030s.
Neural networks, machine and deep learning (ML & DL), speech and text recognition systems
So, we’re getting to the future – neural networks, artificial intelligence, machine revolutions and other horror stories.
Neural networks are perhaps the most interesting technology. With the support of the Internet of Things, 5G and Big Data, it will bring revolutionary changes to our lives.
Additionally, artificial intelligence is any mathematical method that can simulate human intelligence.
Oh as our favorite advertisers and marketers are satisfied… Now any, the simplest neuronetwork can proudly be called «Artificial Intelligence».
However, artificial intelligence is still divided into strong and weak. In 2019, scientists came close to creating a strong AI, the equivalent of human consciousness. This ability not only to distinguish a pen from a pencil or a cat from a dog (according to this principle all neural networks work, it is weak AI), but also to navigate changing conditions, choose specific solutions, model and predict the development of the situation.
A strong AI will be indispensable in intelligent transportation and transportation systems, cognitive assistants. However, this is the future, and what is now?
Now there are learning neural networks. An artificial neural network is a mathematical model modeled on the neural networks that make up the brains of living things. Such systems learn to perform tasks by treating them without specific programming for specific applications. This can be found in Yandex Music, Tesla autopilots, referral systems for doctors and managers.
Therefore, here are the two main trends:
– machine learning (ML – machine learning);
– deep learning (DL – deep learning).
Machine learning is statistical methods that enable computers to improve the quality of the task with experience and training. So it’s about how the neural networks of living organisms work.
Deep learning is not only learning a machine with the help of a person who says what is right and what is not, but also self-learning systems. This is the simultaneous use of different methods of training and data analysis.
However, how do these neural networks teach? What’s the magic?
Actually, in fact, nothing. It’s like training a dog. Neuronetworks show, for example, a picture and say that it is depicted. The neural network must then respond, and if the answer is wrong, it is corrected. An approximate algorithm is given below.
As a result, it turns out that each «neuron» of such a network learns to recognize, refers to it this picture, or rather its part, or not.
Example of neural network operation in image recognition
Neural networks and machine learning apply:
– for forecasting and decision making;
– image recognition and generation, including «pictures» and voice recordings;
– complex data analysis without clear relationships;
– process streamlining.
The application value of this can be seen in the examples of the creation of unmanned cars (decision-making), the search for illegal content (data analysis), the prediction of diseases (pattern recognition and linkage search). At the same time, on the haip it is pattern recognition and generative models (chatGPT, midjourney, etc.). However, business problems are still poorly solved. At the same time, 9 out of 10 students now go to study exactly on pattern recognition and machine vision.
The AI + IoT link deserves special attention:
