A Closer Look at the IoT

There have been experiments with everyday objects such as smart choice, which monitors how much time you sit on it and monitors your vital statistics such as blood pressure, etc. In some of cases, the consumer does not see benefit buying these products.

Not all smart objects provide substantial benefits to the consumer or she may fail to perceive it as a substantial benefit for the extra price. In such cases, the smart product will fail at its launch. Although the technical capability does exist to make the object smarter, the consumer perception does not meet the expectations of the consumer.

There’s also the concept of a smart home, in which objects like dishwashers are controlled out of a central monitoring system. However, due to its expense, this hasn’t gained popularity.

Figure 2-1 illustrates the technological advancement in the field of IoT. It’s reiterated from my book Machine Learning Applications Using Python.

You can see home automation move to the applications stage in the table above because people are currently automating mundane home tasks that they are performing repeatedly, day in and day out, such as switching off appliances and monitoring and controlling them through personal phones like Android and iPhone devices. These applications allow you to control devices like smartwatches, which can accurately gather personal bio information like number of steps taken, heartbeats per minute, blood pressure rate, and oxygen carrying capacity. These apps can generate reports on a weekly, monthly, or yearly basis and use that data to show your health routine.

Another development in the IoT arena is that of the smart city, which is at the stage of early application. The most prominent one is Singapore, which is trying to truly become one in every way. Smart buildings are being built so that they take in maximum sunlight and are self-dependent by using energy recycling methods such as rainwater harvesting solar panels for producing off-grid energy for the residents. A smart city recycles all its waste and consumes and recycles all its resources such as water, oil, garbage, and gas.

Farm automation is one of the examples of an early application of automated operations now possible due to the IoT; however, it has to move to the predictive stage where it is able to produce the crop cycles such as crop failure or crop success. The Weather Company by IBM (www.ibm.com/weather) is trying to change the way farming automation is being done by trying to predict the weather for crops. Today’s farming is completely dependent upon weather cycles and seasons. Although everything in relation to farming, from sowing the seeds to harvesting them, has been automated, the ability to predict the weather for farming will truly move it to the stage of independent operations.

Smart cars are another such application. When you can drive a smart car in crowded areas using IoT sensors such as computer vision sensors and GPS locators, this application will be in the stage of independent operations; also, when it can predict and prevent accidents.

Medical surgery has to a certain extent attained the independent operation level, like medical surgery for cancer where the robot can independently do surgery by augmenting the surgeon’s tasks. 2019 game search applications Other surgeries will mature over time as the world does more research on how to add surgery skills to robots so that they become more efficient.

These are some of the examples of applications of for the IoT to give you a context on how they are placed with technology adoption.

Commercial Uses of the IoT

Let’s now look at some of the successful IoT applications that are working well in the market.

The three solutions that I picked to showcase here work well commercially. The top solution according to the survey is C3 AI Appliance. Appliances that run on different servers are very hard to develop, and it’s harder to meet all customer expectations. Given that this is an AI platform-based appliance running on an Intel server which allows analyzing massive amounts of data and also meets the data governance, compliance, and security protocols, it’s amazing that the customer experience is very high. The customer testimonials on the website are very positive so it’s no wonder it gets a top slot in the survey.

The next product picked from the survey is by Siemens (https://new.siemens.com/global/en/products/software/mindsphere.html). It’s a cloud-based IoT open operating system that connects products, plants, systems, and machines, and lets the customers harness it with advanced analytics. It is an excellent example of an implementation of a platform as a service (PaaS). It’s known as MindSphere. In order to make this another amazing product, given that it is a platform as a service, MindSphere truly works from all popular web browsers such as Chrome, Mozilla, and Firefox, to just name few.

The third product is from all Mbed (https://os.mbed.com/), which created an operating system open source RTOS for IoT, speeding up the creation and deployment of IoT devices based on Arm processors. The main development languages used by this OS are C and C++. It runs on hundreds of hardware platforms such as Raspberry Pi and also has a community that is very active. This platform has various drivers and code examples in libraries available for rapid implementation of all the three functions to compete with each other especially in the MindSphere IoT operating system from Siemens because they are the core for delivering total solutions. C3 AI Appliance takes the concept of an OS further by its ability to host on a private cloud; this solution is boon for those companies that want complete control of their IoT solutions. For companies that want open source solutions, the MindSphere solution by Siemens can work on hardware platforms and specifically uses C and C++ code for software solutions.

You have seen three different companies and three different solutions, and the excellent part is they are delighting customers with their experiences.

IoT Trends for the Future

It is important to look at the top trends that are emerging in the field of IoT in order to get a holistic picture about it. A lot of surveys and prediction of all the IoT trends for the year 2019 have been done. According to ZDNet story on 2019 IoT trends by Eileen Brown (www.zdnet.com/article/whats-next-for-2019-iot-trends-and-predictions/) a Northstar survey of global consumers showed that intelligent homes will become mainstream and the next hot thing will be interdelivery options to consumers that will be delivered through smartphones and GPS positioning data. It also predicted better quality in healthcare due to the deployment of connected sensors in hospitals to ensure that the time to find critical medical equipment is reduced. The survey also talked about the function of smart cities by delivery cost reduction benefits in terms of better waste management and citizen engagement for revenue stream opportunities and energy efficient buildings. This has been in the mainstream now with cities like Singapore, which has already shown that a smart building can reuse its waste including water and garbage resources to be self-sufficient.

A survey by GoodWorkLabs on the IoT, published in February 2019 (www.goodworklabs.com/iot-trends-2019/), talked about edge computing, which is distributed computing performed on distributed smart edge devices instead of in a centralized environment for indoor areas or industries where there is no requirement for a centralized network for processing. The survey lists as an important item security for IoT devices and notes the efforts taken by hardware manufacturers like Siemens and GE, which are making smart devices that focus on the endpoint security of the user. The survey also talks about its application in the healthcare and manufacturing industries. The smart beacon RFID tags are part of the new industrial revolution that is going to take place on devices by 2020. The governance of consumer IoT industry is also pointed out to be a trend that will advance the development of smart homes. Another major development is that large players in the consumer industry are coming together to form subscription offerings. This will happen in the areas of utilities, food companies, and service aggregators. There are talks about the growing market of connected smart cars which will be accessible through the smart apps on your mobile phones, which will show you real-time diagnostic information about your car using IoT technology. All the sensors residing within your car will give you data about your current location but also what is going in your car such as tire pressure. The survey welcomes the beginning of 5G in 2019, which will become the backbone of IoT technology by supporting the interconnectivity of IoT devices. The 5G network will allow smart devices to produce and send data in real time, which has not been imagined so far.

The website iotforall.com (www.iotforall.com/top-iot-trends-rule-2019/) reports that 2019 will be about big data and artificial intelligence. The topmost forecast by Gartner says that there will be a rise of 14.2 billion IoT deployments. It talks about connected clouds. Many companies will rely on clouds to store data because cloud storage is connected to bandwidth. Accessing data is the reason for choosing different cloud services, such as the offerings by giants like Amazon AWS Cloud Platform, Microsoft Azure, and the Google Cloud platform. Today’s need is for these different clouds to speak to each other; this is known as connected clouds, such as the partnership formed by Oracle and Microsoft by connecting their cloud platform services. The survey also talks about connecting private and public cloud for any company for servicing their data storage needs. The survey also talks about edge computing, which is going to break through in 2019. The survey talks about digital twins, which are also known as hybrid or virtual prototyping businesses using special tools like AI and machine learning and IoT to improve their customer business experiences by streamlining their data operations. This survey highlights that 5G will be game-changing in the IoT market by bringing about revolutionary applications. One point mentioned by this survey is sensor innovation, although it does not talk about what type of sensors will be developed in the market, but it says that the new special purpose sensors will lead to efficient and effective use of power consumption using deep neural networks, leading to new age architectures and low power IoT and endpoint devices. The development of new algorithms using data from these sensor technologies will lead to new implementations in the domains where they are applied. The survey prediction for social IoT, which is giving rise to social anxiety and is going to transfer the business sector from consumer devices to large scale manufacturing, is like a warning for the IoT applications and solutions implementers in order to take social responsibility into considerations when building their products. The social aspect is highlighting the need for a human being to accept the current state of human beings in accepting IoT in their everyday life. Although the survey does not point to any specific issues, it says that as online gains acceptance, social, legal, and ethical issues will crop up.

A survey by Network World covers the top 10 IoT trends for 2019 and beyond (www.networkworld.com/article/3322517/a-critical-look-at-gartners-top-10-iot-trends.html). The first trend is artificial intelligence, which is not a big surprise. The second one is social, legal, and ethical issues. Info Linux and data booking which river is commercializing of Information and making the data available agents and submergence on subscription or condition basis which this Gartner study reports. IV is shift phone edge intelligent mesh. The edge mesh defined as a network of connected edge devices true internet connect with cloud end-user devices together at a high level known as edge mesh. The study points to the fact that instead of having such devices working in silos, this will get them connected to intelligent machines and will create a network of networks of their own. The trend, as seen by Gartner, is that of IoT governance, covering the legal and social aspects and ethics involved in the implementation of IoT applications.

The other oils like new user experiences and new Wireless Networking Technologies for IoT salwar significant thanks that the study mentions.

PCMag (www.pcmag.com/feature/365945/8-iot-trends-to-watch-in-2019/8) came out with its trends to watch in 2019. One issue that the IoT is currently facing is surcharge protection against ransomware. It mentions the case of the city of Atlanta, when city systems such as the water services system were hacked. by a hacker it systems ransom technicians that all the smart devices this threat mobile device Management Solutions have to give up in order to protect them against search attacks. An interesting case has been highlighted survey date of IoT Technology tablets used to keep food safer. Example of cold food storage operator which offers hybrid cloud to keep food safe IoT is helping companies teacher and humidity cold storage facilities.

Another issue highlighted through this article is the use of IoT to simplify maintenance in manufacturing. The use of sensors to gauge problems in manufacturing has been discussed where the actual site technician only thumbs the sensors on the touch screen to detect problems that could arise with the machine in the near future. Factory workers are going to have IoT-based wearables on their bodies to manage, monitor, and control them from control centers, as highlighted in this article.

A TechGenix article on trends and predictions for 2019 talks about two interesting aspects of the new IoT platform (http://techgenix.com/iot-trends-2019/). It talks about IoT platform vendors that are not focusing on use cases and integrity. The hyperscale clouds are going to be labeled as IoT destinations; a marriage between hyperscale cloud IoT platforms is going to emerge from this year onwards. Another trend is the development of the managed services market for the IoT. The emergence of IoT services will include management operation of IoT networks and other IoT assets. In the article, the prediction is that these offerings will be customized for smart products made for smart homes to support IT solutions.

The last survey is the article by Forbes (www.forbes.com/sites/bernardmarr/2019/02/04/5-internet-of-things-trends-everyone-should-know-about/#752315be4b1f) written by Bernard Marr in February 2019. Among other things, it predicts that devices will become more vocal. For example, Alexa, Siri, and Google Voice commands are going to see their emergence.

After having looked at the common future trends and predictions in IoT from the best in the industry, I now present to you a trend that is definitely going to see its emergence starting in 2020. The Thought AI has yet to see its commercial implementation, including inventions of thought sensors. I will discuss all of this in the next section of this chapter.

A Closer Look at Thought AI

You have seen the best IoT commercial applications in the field of IoT. You have also seen the trends for the future as predicted by the aforementioned articles. However, as a technical visionary, I do not think that any of these trends are going to be as game-changing as a certain application. Yes, I am talking about something that is going to not only revolutionize but also change the way humans and machines interact with each other. It is going to create a world that is going to be more intuitive, easy going, and responsive to human needs. It will bring about a change as revolutionary as the invention of the PC or the mobile phone. It will see technology move up to a level that has not been seen before. It will make technology rise extremely close to the Homo sapiens as could ever be. It will be the age of machines as friends and rogues. Of course, such technology has the equal potential of being anti-human as well. I re-quote from my first book, Machine Learning Applications Using Python, “It is not the artificial intelligence but the human intelligence behind the artificial intelligence that is going to change the way we live our lives in the future”. Please remember that this section is futuristic and that Thought AI is a basis for the near-future as it is yet realized. It can happen in the future. Also, you need to know that implementing Thought AI will be costly as it is in the research and development stage.

As a visionary, I present to you the technology for tomorrow. This is what I call Thought AI. If you look at the input methods that exist in any machine from a human being, there are just two. The first one is typewriting via a keyboard and the second one is voice dictation via a speaker. Table 2-4 lists the differences between the type and voice inputs to emulate machine-like robots on our computer.

Figure 2-2 shows the process that the human brain has to undergo to produce a typewritten work. You need to read them together to understand the advantages of Thought AI.

Table 2-4

Advantage of Using Thought AI

Typewriting	Voice Dictation	Thought AI
Advantages
It is an old way.	It is a new way.	It is going to be the new way.
Many people know how to type on a keyboard.	Everyone who can speak can work with this input.	Everyone who can think can work with this input. Even persons with disability can use it.
Keyboards in new devices are now built into the software. No external device is necessary to type.	Microphones are built into every device these days including PCs, laptops, mobiles, and tablets.	Commercial grade thought sensors have yet to be developed. Once in the market, they will be built into every device we know of today such as mobiles, PCs, laptops, etc.
Private activity as long as the screen is not visible to others.	Not as private because what you speak is audible to everyone around you.	Totally private to you. Nobody except the device that is running the Though IoT sensor knows what you are thinking.
Disadvantages
Takes time	Takes time but it’s faster than typing.	Advantage: Real-time input as the term will get popular “WHAT YOU THINK IS WHAT YOU GET.”
Requires special practice to type fast and efficiently.	Does not work in noisy environments.	Advantage: Does not require any special skill like learning to bang a keyboard. You just think.
Tiresome activity if pages of input are needed. Hands get tired of typing.	Speaking continuously for large input can be tiring.	Advantage: You get tired only if you get tired of your thoughts.
Not real-time to our brain activity. There’s a lag between our thought flow and typing activity on the keyboard. The brain has to do extra processing to convert thoughts to the motor activity of spelling each word and then giving each finger the command to type.	Lack of privacy because the voice is audible to everyone around the human being.	Advantage: Can’t get more real-time because the inputs are received via the thought sensor as and when you think.
	Challenges of different accents around the world for just the English language.	Advantage: No issue with accents because thoughts are universal in a particular language.

Industrial Internet of Things

It is now time to understand the Industrial Internet of Things (IIoT). The IIoT plays the industrial equivalent of the IoT; it is computers, devices, and objects that may be connected to each other and share shop data with each other. The collected data resides as a central service like the cloud-based service or resides in private computers which they access for their own use. IIoT works in the manufacturing world. IIoT allows access to industrial data at better speeds and direct use of connected devices in the factory environment. Many IIoT protocols such as the Modbus or MQTT protocols allow IoT devices to directly link with industrial devices like assembly-line machines, processing machines, boilers, or heat exchangers, to just name a few. So what is the advantage of connecting today’s factory machines, which are already monitored by their own software of some kind? Yes, it is true that any kind of an industrial machine that is built today has some kind of monitoring and control system which has data that is available from within the machine since it has a limited storage capacity recycled for continuous operation. Let’s consider a heat exchanger, as part of a power plant. It has data about temperature details like heat and pressure within the control panel. However, if we want to make the single unit machine more efficient and productive, there is some kind of a pattern operating them and this data can be centralized on a public or private cloud. You could also apply some IoT sensors to the heat exchanger in order to collect different data that is not generated within it. So we see in this simple example that the biggest benefit can be cost reduction if we are able to collect data from machines currently sitting in silos and then apply learning and deep learning on them to gain an advantage. This can be seen from Figure 2-3.

In the diagram, there is an industrial machine with its own control panel. (It does not really matter what kind of industrial machine process; I am showing you that it’s universally applicable.) There all are four major IoT systems. The first one is the industrial machine itself. The second is its hello. The third is the private cloud. The fourth are the IoT sensor devices that affect the industrial machine. V continent really is the heart of IoT system which provides the real benefit behind machine learning or deep learning model that is built using the private or public cloud. Industrial machines cannot communicate directly with IoT sensor devices, so they have a control panel that faces into the parts of the machine to monitor and create data. We generate machine-level internal data access through the control unit interfacing a device using the MQTT protocol. These common industrial communication protocols help us access the private data and send it all to the private cloud. If we were to monitor the effect of temperature on the industrial machine, we would have to install IoT sensor devices on the industrial machine. After installing the IoT sensor device, this device would send data to the private cloud and integrate it into a centralized database. Once we have enough data, we could apply machine learning or deep learning models to this data. It is now up to the business to check the results of machine learning or deep learning model and to apply efficiencies on the industrial machines. So with this example, you should now have a clear picture of how machine learning or deep learning is applied in an industrial environment. IIoT applications can include energy audits, machine efficiency checks, and improving factory efficiencies. The applications are limitless. Now let’s look at some successful IIoT applications that are working well in the market.

IoT and IIoT Differences

The focus of IoT applications is the consumer at large and the focus of IIoT is industrial applications. This actually determines the ground for the next two parameters: accuracy/precision and risk impact. The accuracy and precision of industrial grade applications should be higher because they sometimes deal with hazardous processes and impact many lives on the factory shop floor. An error by an IIoT application could cause a company millions or even billions of dollars of losses and endanger the human lives involved in the production processes. This is the reason why IIoT devices use sensors that have higher accuracy levels than those used in personal IoT applications. The risk impact of an IIoT application failure is very high because many lives and a lot of business money and resources are at stake. IoT applications are also used in critical processes like healthcare where lives could be at danger; however, they are of personal nature. Nevertheless, every life is important and has value. With this, we come to the end of this chapter. Next you’ll see how machine learning is used with IoT and IIoT in Python with example code.

Summary

In this chapter, you learned what IoT and IIoT are all about. You now understand the differences in their implementation based on focus, accuracy, and risk impact. You looked at some commercial IoT applications such as C3 AI Appliance, MindSphere (the cloud-based IoT open operating system from Siemens), and Mbed OS (the leading open-source RTOS for the Internet of Things). Next, you looked at the transformation of AI for human needs from typewriting to dictation to Thought AI, which is in the future. You looked in depth into IIoT and how industrial communication protocols work together for industrial applications. You explored the commercial use of IIoT by companies such as Corlina, which certifies edge devices and the General Electric Predix platform for IIoT devices and solutions, to name a few. Lastly, you learned the differences between IoT and IIoT in order to understand how they are implemented in real-world environments.