Scribd Logo
Upload

Welcome to Scribd!

  • Iot Notes

    Uploaded by

    Er.Utkarsh Pandey
    11 views33 pages

    Original Title

    IOT NOTES

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    11 views33 pages

    Iot Notes

    Uploaded by

    Er.Utkarsh Pandey

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    of 33
    INTRODUCTION TO INTERNET OF THINGS (10T) internet of Things (IoT) rhe Internet is a powerful tool used in all kinds of the information systems today. The network is available almost anywhere, at home, at work and also on mobile devices (phones, watches). People tart to think to connect the Internet to almost all devices of everyday use, s° they can communicate stn each other by taking simple decisions for people and helping them in their life, Such idea is viiled the Internet of Things (IoT), It is estimated that currently about 15 billion devices are NORTH Intemet of Things and Artic inttigencg 4 + Gateway and Network Layer The data coming out from the sensors and the connectivity network layer is to be routed to the Management service layer through the Gateway and Network layer. In order to collect the enormous. amount of data from the sensors and RFID tags, this layer needs to have a large storage capacity, Bach IoT device works on different kinds of network protocol. So this layer should be responsible fa, integrating all the network protocols. Examples for gateway networks are LAN (Local Area Network), WAN (Wide Area Network) etc. : + Management service Layer The key role of this layer is to manage all the IoT devices. This layer consists of information analysis device management ete. Sensors collect enormous amount of raw data. Data management incleda, extracting necessary information from the data collected by the sensor and to provide a valuabl, result of all the data collected and this is the main responsibility of Management Service level + Application Layer The top most layer of IoT architecture is Application Layer. Application layer is responsible for effective utilization of collected data. Different applications of Io are Home automation, Industrial Automation, health monitoring and so on. Internet of Things (IOT ) Communication Technologies / Protocols: Based upon covered course till now, we are familiar with IOT basics , IOT working , IOT application & 1OT architecture. Now we are going to cover deeper understanding topics of IOT Communication Technologies and protocols, which will illustrate the communication technologies & protocol standards which are used by IOT devices to connect & interact each other. IOT Technologies and Protocols > Fig, 1.19: IOT Communication Technologies Several Communication Protocols and Technology are being used in the internet of Things. Some of the major IoT technology and protocol (oT Communication Protocols) are Bluetooth, Wifi, Radio Protocols, LTE-A, and WiFi-Direct. These IoT’ communication protocols cater to and meet the specific functional requirement of an Io system. Majorly used 7 important I0T communication technologies are described below. ipnosuctiont0 Internet of Things (IOT) 15 Bluetooth: n important short-range IoT communication Protocol / Technology, Bluetooth which has become very important in computing and many consumer product markets. It is expected to be key for wearable products in particular, again connecting to the IoT albeit probably via a smartphone in many cases. ‘The new Bluetooth Low-Energy (BLE) - or Bluetooth ‘Smart, as it is now branded — is a significant protocol for ToT applications. Importantly, while it offers a similar range to Bluetooth with significantly reduced power consumption, Bluetooth Role in The Future of loT Fig 1.20: OT Communication Technologies-Bluetooth b.Zigbee : ZigBee is similar to Bluetooth and is majorly used in industrial settings. It has some significant, advantages in complex systems offering low-power operation, high security, robustness and is well positioned to take advantage of wireless control and sensor networks in loT applications. The latest version of ZigBee is the recently launched 3.0, which is essentially the unification of the various ZigBee wireless standards into a single standard. e® @ © Fig. 1.21: OT Communication Technologies-Zighee ©. ZWave: Z-Wave is a low-power RF communications IoT technology that primarily design for home automation for products such as lamp controllers and sensors among many other devices. A Z-Wave uses a simpler Protocol than some others, which can enable faster and simpler development but the only maker of ee EE a NORTH Internet of Things and Artificial ntetigence chips is Sigma Designs compared to multiple sources for other wireless technologies such as ZigBee and others. Cawove " hk Fig 1.22: 1OT Communication Technologies-Z-Wave d. Wi-Fi: WiFi connectivity is one of the most popular IoT communication protocol, often an obvious choice for many developers, especially given the availability of WiFi within the home environment within LANs. There is a wide existing infrastructure as well as offering fast data transfer and the ability to handle high quantities of data. Currently, the most common WiFi standard used in homes and many businesses is 802.11n, which offers range of hundreds of megabit per second, which is fine for file transfers but may be too power-consuming for many IoT applications. e. Cellular: Any IoT application that requires operation over longer distances can take advantage of GSM/3G/4G cellular communication capabilities. While cellular is clearly capable of sending high quantities of data, especially for 4G, the cost and also power consumption will be too high for many applications. But it can be ideal for sensor-based low-bandwidth-data projects that will send very low amounts of data over the Internet. Fig. 1.23: |OT Communication Technologies-Cellular introduction to Internet of Things (IOT) £.NFC: FC (Near Field Communic: wiveen i ay is an IoT technology. It enables simple and safe communications mernsactions in which ore dace oy eteally for smartphones, allowing consumers to perform arent and connest chet ent have to he physically present Ie helps the user to access digital trannlogy and enables devi coe ewce® Essentially it extends the capability of contactless card share information at a distance that is less than 4em "7 Je y = wp Ientfcstion Cauhleas Payment Fig, 1.24: IOT Communication Technologies-NFC g. LoRaWAN: LoRaWAN is one of popular IoT Technology, which targets wide-area network (WAN) —_ applications. The LoRaWAN design to provide —= low-power WANs with features specifically C communication in IoT, smart city, and industrial — applications. Specifically meets requirements for = low-power consumption and supports large networks with millions and millions of devices, Fig: 1-25: 1OT Communication Technologies-LoRa-WAN data rates range from 0.3 kbps to 50 kbps. So, this was all about IoT Communication Technology. Internet of Things (IOT) protocols A protocol is a standard set of rules that allow electronic devices to communicate with each other Communication protocol is a defined set of rules and regulations that determine how data is transmitted in telecommunications and computer networking oa NORTH Internet of Things and Artic Inteligg noe 0 ‘The IOT Protocols are the Communication Protocols, which are being used in 1OT ; i e Majorly Used 4 Key IOT protocols are detailed here. ‘ommunications 1OoT Protocol is Fig 1.26: 1OT Protocols 1. Constrained Application Protocol (CoAP) : CoAP is an internet utility protocol for constrained gadgets. Itis designed to enable simple constrained devices to join ToT through constrained networks having low bandwidth availability. This protocol is primarily used for machine-to-machine (M2M) communication and is particularly designed for IoT systems that are based on HTTP protocols. f ig. 1.27: CoAP makes use of the UDP protocol for lightweight implementation. It also uses restful architecture, which is just like the HTTP protocol. It makes use of dtls for the cozy switch of statistics within the slipping layer. Fig. 1.28: 1OT MQTT Protocol inroduction to Internet of Things (IOT) 19 2-Message Queue Telemetry Transport Protocol (MQT': Me Cleat ct Telemetry Transport is a messaging protacol developed with the aid of Andy Stanford lark of IBM and Arlen Nipper of Arcom in 1999 and is designed for M2M communication Us normally {04 for far away tracking in tT. It primary challenge isto gather statistics from many gadgets and delivery of its infrastructure MQTT connects gadgets and networks with packages and middleware, AM the devices hook up with facts concentrator servers like IBM's new message sight aoplianee, MTT protocols paintings on top of TCP to wlfer easy and dependable. streams of information. These ToT protocols include 3 foremost additives: subseriber, generates the information and transmits the facts to subserib guarantees safety by means of move-checking the authorization of publisher, and dealer. The writer ers through the dealer. The dealer f publishers and subscribers. 3, Advanced Message Queuing Protocol (AMQP): This was evolved by John O'Hara at JP Morgan Chase in London. AM@P is a software layer protocol for message-oriented middleware environment, It supports reliable verbal exchange through message transport warranty primitives like at-most-once, at least once and exactly as soon as shipping. The AMQP — IoT protocols consist of hard and fa ist components that route and save messages within a broker carrier, with a set of policies for wiring the components together. The AMQP protocol enables patron programs to talk to the dealer and engage with the AMQP model. This version has the following three additives, which might link into processing chains in the server to create the favoured capabilities. + Exchange: Receives messages from publisher primarily based programs and routes them to ‘message queues’. + Message Queue: Stores messages until they may thoroughly process via the eating client software + Binding: States the connection between the message queue and the change, Fig, 1.29: IOT AMQP Protocol ae oo NORTH Internet of Things and ArtifiaIneigengg 4, Data Distribution Service (DDS) : It enables a scalable, real-time, reliable, excessive-overall performance and interoperable stat change via the submit-subscribe technique. DDS makes use of multicasting to convey high-qu, QoS to applications. ists ality DDS is deployed in platforms ranging from low-footprint devices to the cloud and supports gree, bandwidth usage in addition to the agile orchestration of system additives. ‘The DDS — Io protocols have fundamental layers: facts centric submit-subseribe (deps) and statistic, local reconstruction layer (dlr), Deps plays the task of handing over the facts to subscribers and th, irl layer presents an interface to deps functionalities, permitting the sharing of distributed dat, amongst IoT enabled objects. Application 1 Application 2 Application 3 | Fig, 1.30; JOT AMQP DDS Protocol Above mentioned were the 4 important IoP protocols to make thorough with this IOT Protocol concept Internet of Things (IOT) Characteristics IoT is a global infrastructure for information, enabling advanced services by interconnecting physits and virtual things based on existing and evolving information and communication technologies. lol represents a convergence of several domains and can be perceived as an umbrella term. ‘The added value to the businesses through IoT is created by the information that is collected by lol devices which go through five phases of IoT lifecycle + Firstly, ereate phase, where devices or sensors collect information from the physic environment around them. 'The data from smart connected devices can be used to genert® insights that can help businesses, customers and partners. Introduction to Internet of Things (107) a «secondly, communicate network to the desired d + thirdly, Phase, where the data and events generated are sent through the lestination, 0&Bregate phase, where data collected are aggregated by devices itself. + fourthly, analyse phase, where, upon further sophisticated analytics the aggregated data can be used to generate basic Patterns, control and optimise processes. * finally, act phase, where suitable actions are performed based on the information eveated The ToT is @ complex system with a number of characteristics. Ite characteristics vary from one domain to another. Some of the "ie general and key characteristics identified during the research study are as follows: 1. Intelligence Jo? comes with the combination of algorithms and computation, software & hardware that makes it smart, Ambient intelligence in IoT enhances its capabilities which facilitate the things to respond in an intelligent way to a particular situation and supports them in carrying out specific tasks. In spite of all the popularity of smart technologies, intelligence in IoT is only concerned as means of interaction between devices, while user and device interaction is achieved by standard input methods and graphical user interface. 2. Connectivity Connectivity empowers Internet of Things by bringing together everyday objects. Connectivity of these objects is pivotal because simple object level interactions contribute towards collective intelligence in IoT network. It enables network accessibility and compatibility in the things. With this connectivity, new market opportunities for Internet of things can be created by the networking of smart things and applications, 3. Dynamic Nature ‘The primary activity of Internet of Things is to collect data from its environment, this is achieved with the dynamic changes that take place around the devices. The state of these devices change dynamically, example sleeping and waking up, connected andior disconnected as well as the context of devices including temperature, location and speed. In addition to the state of the device, the number of devices also changes dynamically with a person, place and time. 4. Enormous scale ‘The number of devices that need to be managed and that communicate with each other will be much larger than the devices connected to the current Internet. The management of data generated from these devices and their interpretation for application purposes becomes more critical, Gartner (2015) confirms the enormous scale of IoTT in the estimated report where it stated that 5.5 million new things will get connected every day and 6.4 billion connected things will be in use worldwide in 2016, which is up by 30 percent from 2015. The report also forecasts that the number of connected devices will reach 20.8 billion by 2020. INTRODUCTION TO ARTIFICIAL INTELLIGENCE (@il), MACHINE LE@RNING (ML) AND DEEP LEARNING (DL) Artificial intelligence, Machine Learning and Deep Learning are the trending topics worldwide with the applications being deployed in all the major social & business segments. Artificial Intelligence is a broader umbrella under which Machine Learning (ML) and Deep Learning (DL) comes. Diagram shows, ML is subset of Al and DL is subset of ML. STOTT MHS ig Peron tere te oer La a nore DEEP LEARNING’ © Cree te) array posuere a Fig 4.1 Artificial Intellegence , Macl \¢ Learning & Deep Learning Artificial Intelligence: Al is composed of 2 words Artificial and intelligence. Anything which is not natural and created by humans is artificial. Intelligence means ability to understand, reason, plan ete. So we can say that any code, tech or algorithm that enable machine to mimic, develop or demonstrate the human cognition o” behaviour is AI. “The study of the modelling of human mental functions by computer programs.” — Collins Dictionary ‘The concept of Al is very old but it got popularity recently. The reason being earlier we had a very small amount of data to make accurate predictions. But today, there is tremendous increase in the siz@ of the data which is generated every minute and help us to make more accurate predictions, Along with the enormous amount of data, we also have the support of more advanced algorithms, high en! computing power and storage that can deal with that huge data size. Examples include Tesla set driving car, Apple's Siri and many more 76 > introduction 10 Artificial Intelligence (Al), Machine Learning (ML) & Deep Learning (DL) 7 Machine Learning : pue to enormous Complex Statistical Calculations & Robust AI versions requirement, — Machine yearning was introduced. sMachine learning is the science of getting computers to act without being explicitly programmed. — Stanford University Machine Learning is a subset of AI which uses statistical methods to enable machines to improve with experience. It enables a computer to act and take data driven decisions to carry out a certain task These programs or algorithms are designed in such a way that they can learn and improve over time when exposed to new data, Example: Suppose we want to create a system that tells us the expected weight of person based on its height. Firstly, we will collect the data. This is how data looks like (picture below). Each point on graph represents a data point. | e e e.°* oe e x eign em To start with, we will draw a simple line to predict weight based on height. A simple line could be W = H-100 Where W=Weight in kgs H=Height in ems eight in ko Heightin em eN 78 INORTH Internet of Things and Artificial inteligencg ‘This line can help us to make prediction. Our main goal is to reduce distance between estimated value and actual value. i.e the error. In order to achieve this, will draw a straight line which fits through a) the points. Weight in ko elght nem Our main goal is to minimize the error and make them as small as possible. Decreasing the error between actual and estimated value improves the performance of model and also the more data points we collect the better our model will become. So when we feed new data that is height of a person then it could easily tell us the weight of the person, Deep Learning: “Deep Learning is a subfield of machine learning concerned with algorithms inspired by the structure and function of the brain called artificial neural networks”. — Machine Learning ‘Masterylt’s a particular kind of machine learning that is inspired by the functionality of our brain cells called neurons which lead to the concept of artificial neural network (ANN). ANN is modelled using layers of artificial neurons or computational units to receive input and apply an activation function along with threshold. In simple model the first layer is input layer, followed by a hidden layer, and lastly by an output layer Each layer contains one or more neurons. Hidden Inpat Output SNS icial neural Network Introduction to Artificial Inteligence (Al), Machine Learning (ML) & Deep Learning (DL) 79 Simple Example to understand how things happen at conceptual level for ML & DL. How you recognize square from other shapes? First thing we do is check whether the figure has four lines. If yes, we further check if all are lines are connected and closed, If yes we finally check if all are perpendicular and all sides are equal. We consider the figure as square if it satisfies all the conditions. ‘As we saw in the example it’s nothing but nested hierarchy of concepts. So we took a complex tas sk of identifying a square and broken down into simpler tasks. Deep learning also does the same thing but at a larger scale. For instance, A machine performs a task of identifying an animal. Task of the machine is to identify weather given image is of cat or dog. IS THIS A CAT «DOG? | AF 006: ‘activate Mvnens occe Wien gumiad aS ecm Differentiating between dog & cat using Deep Learning algorithm If we would have asked us to solve this using concept machine learning then we would have defined features such as check if it has whiskers or not, check if tail is straight or curve and many other ures, We will define all features and let our system identify which features are more important in sifying a particular animal. Now when it comes to deep learning it takes it to one step ahead. Deep faa INORTH Intemet of Things and Artificial Intelligence learning automatically finds which features are most important for classifying as compared to machine learning where we had to manually give out the features. So till now we understood AI is bigger picture and machine learning & deep learning are its subparts of it, Machine learning (ML) vs Deep learning (DL) Easiest way to understand difference between machine learning and deep learning is “DL IS ML” More specifically it’s the next evolution of machine learning. Let's take few important parameters and compare machine learning with deep learning, 1. Data Dependency ‘The most important difference between the two is the performance as the data size increases. From the graph below we can see that as the size of the data is small deep learning doesn't performs that well but why? ‘This is because deep learning algorithm requires large amount of data to understand it perfectly. On the other hand machine learning works perfectly on smaller datasets. nt of data 2. Hardware Dependency spendent on high end machines while machine learning well. This is because requirement of deep learning art of its working, GPU's are required as they perform rnd these operations are only be efficiently optimized Deep learning algorithms are highly de algorithms can work on low end machines as algorithms include GPU's which is an integral p: large amount of matrix multiplication operations at if they use GPU's. 3. Feature engineering Ws the process of putting domain knowledge to reduce the complexity of data and make patterns mel visible to learning algorithms. This process it’s difficult and expensive in terms of time and expertise In case of machine learning, most of the features are to need be identified by an expert and then hand coded as per the domain and data type. The performance of machine learning depends upon how accurately features are identified and extracted. But in deep learning it tries to learn high level. features from the data and because of this it makes ahead of machine learning. i - Introduction to Artificial Intelligence (Al), Machine Learning (ML) & Deep Learning (DL) 81 4, Problem Solving Approach When we solve problem using machine learning, its recommended that break down the problem into sub parts first, solve them individually and then combine them to get the final result, On the other hand in deep learning it solves the problem end to end, For instance, the task is multiple object detection ie what is the object and where it is present in the image. 0 let's see how this problem is tackled using machine learning and deep learning. In a machine learning approach, we will divide problem in 2 parts: object detection and object recognition. We will use an algorithm like bounding box detection as an example to scan through image and detect all objects then use object recognition algorithm to recognize relevant objects. When we combine results of both the algorithms we will get the final result that what is the object and where it is present in the image In deep learning it perform the process from end to end. We will pass an image to an algorithm and our algorithm will give out the location along with the name of the object. 5. Execution Time Deep learning algorithms take a lot of time to train. This is because there are so many parameters in a deep learning algorithm that takes the training longer than usual. Whereas in machine learning the training time is relatively less as compared to deep learning. Now the execution time is completely reverse when it comes to the testing of data. During testing deep learning algorithms takes very less time to run whereas the machine learning algorithms like KNN test time increases as the size of the data increases. 6. Interpretability This is the main reason why people think a lot before using it in the industry. Suppose we use deep learning to give automated essay scoring. The Performance it gives is excellent and same as human beings but there are some issues that it doesn't tell us why it has given that score, indeed mathematically it’s possible to find out which nodes of deep neural network were activated at that time but we don't know what the neurons were supposed to model and what these layers were doing } Lap, INORTH Internet of Things and Artificial Inteligence a colleetively. So we fail to interpret the result but in machine learning algorithms like decision tree eves us a crisp rule that why it chose what it chose so itis easy to interpret reasoning behind it With above study, itis clear the Understanding of AI, ML & DL. and what relationship they share and how they are different from each other. ‘Artificial intelligence is the best solution to manage huge data flows and storage in the IoT network. IoT now a days becoming more and more popular with the inventions of high-speed internet networks and many advanced sensors that can be integrated into a microcontroller. The data flows internets now will have sensors data and user data that send and receive from the workstations. With the | Intelligence (Al) in Internet of Things (1OT) Why IoT and Al need each other? Senses Ga = ey Decides ‘Remember The use of 1o7, AJM and Big Data analytics play an important ote in making the elt effictent. increase in the number of workstation and more and more sensors, some data may be facing problems on the storage, delay, channels limitation and congestion in the networks. To avoid all these problems, there were many algorithms were proposed in the past of 10 years. Among alll the algorithms, Artificial Intelligence still being the best solution to the data mining, manage and control of congestion in the network. The aim of this paper is to present the application of artificial intelligence system in the IoT. Also, the method used in the Artificial Intelligence like fuzzy logics and neural network also will be discussed in this paper in conjunction with IoT network. The self: optimizing network and software defined network are parts of the important parameters in the Artificial Intelligence IOT System. Applications of Artificial Intelligence (Al) in Internet of Things (1OT) ‘Technologies like Internet of Things (IoT) and Artificial Intelligence (Al) tell us that the future is now. Remarkably, these technology concepts perfectly complement each other. The number of connected devices will only expand and the mass of data produced by them will grow to head-spinning volumes. Al can help organizations gain meaningful insights from big data that loT provides. ‘AL & IOT are being used jointly to make all services and Industry more convenient & effective with better result. Below diagram shows the all applications Sectors. F Introduction 10 Artificial Intelligence (Al), Machine. Learning (ML) & Deep Learning (DL) 83 Transport & a = aa Logistica Utilities Smart cities | Smart building Fioet management, | Smartmetering Parking sensors, ‘Smoke detector, Goods tracking | Smart rid management] Waste management, ete. | Home automation — Consumers Industrial Environment Agriculture | @ Z ~] 7 le ey | o- Process monitoring & i 10 Wearables cont. | Focd montoringatens, | “maneaicel Kidsisenior tracker | ainance montering | E*vionmental monitoring | ivestock vac Examples of AI Applications in oT Let's take a closer look at industries and businesses that have already managed to cut costs, create a better user experience and open up new business models with AI in IoT. Perhaps, these examples will make you consider implementing AI and IoT in your business. 1, Industrial Internet of Things (IIoT) The widespread use of IoT devices with industrial equipment provides a plethora of data. With Al algorithms applied to the gathered data, business owners can detect potential issues, fix them in advance and apply these insights to other cases. The system is gradually taught to recognize external and internal factors that have an impact on the operation of the machines. By optimizing resources and increasing industrial safety, the entire production process is streamlined. Predictive maintenance is the brightest showcase of AI used in IloT. Predictive or perspective maintenance means that a system powered by the machine learning algorithms can predict a need for maintenance on a plant floor. On top of that, artificial intelligence can help in creating self-healing and self-calibrating IoT devices like sensors, inductors or transmitters. The biggest benefits that AI adds to the IIoT due to the described capacities are the reduction of maintenance costs and downtime. 2, Healthcare ‘The healthcare industry generates showers of data. Sensors from medical devices, healthcare mobile apps, fitness trackers and digital medical records have been producing and collecting patients data for years. The AI and IoT approach can help predict diseases, suggest preventive maintenance and provide drug administration. When it comes to health protection or disease control, patients and hospitals would welcome the benefits that come with the AI and IoT approach. iT “ INORTH Internet of Things and Artfcaincigence 3. Smart Home ‘The idea of a fridge “communicating” with a smart watch is still just a concept. Even so, “intelligent” vacuum cleaners, doorbells and lightning systems are found on the market in good supply. According, to IDC, consumers will make more investments in smart home ecosystems to the tune of $63 billion by 2020. Will there be a place for artificial intelligence at home which is “smart” enough already? We bet it will. Artificial intelligence means even bigger automation in a smart home. Since the entire idea of connected objects is to make a life easier, more automation sounds great. On top of that, Al can make life in smart homes even more pleasant. AI systems can “learn” your mood and preferences, as well as analyze your interaction with home objects. With such knowledge, it can adjust temperature for both heating and cooling, adjust lighting, put on the music you like and close or open windows depending on the weather. IoT and machine learning can also water plants when sensors state the dry soil and start a vacuum cleaner every Saturday at 1 PM. 4, Business Growth with More Security AI and Machine Learning as its subfield are a huge leap for businesses that make active use of IoT. Despite certain concerns largely connected with safety and security of AI in IoT, a combination of these disruptive technologies has already been successfully tried out. Business goals become more attainable with actions taken in advance. The ability to analyze, predict and automatically adjust to a particular need is highly prized. When it comes to AI and IoT, global tech leaders have already learned some ropes. Like Google, ‘Amazon, Tesla, Uber and other tech giants go on exploring intelligent tech, one thing is certain — the impact of AI and IoT is going to be tremendous. 5. Self-driving vehicles ‘Automated Self driven cars such as that of Tesla Motors is a good example of Al application in IOT , where 6. Automated vacuum cleaners Automatic Vacuum cleaner having use of Robot ie. iRobot Roomba is a great example of Al application in IOT. 7. Smart thermostat solutions Automatic Thermostat Solution by Nest Labs for Automatic AC operation & temp Control is also 4 good Example of AI application in IOT. When it comes to AI and IoT, global tech leaders have already learned some ropes. Like Google, ‘Amazon, Tesla, Uber and other tech giants go on exploring intelligent tech, one thing is certain —the impact of AI and IoT is going to be tremendous. vest Arica Ineligence (Al), Machine Learning (ML) & Deep Learning (OL) 85 managing, and Analysis data generated by IOT devices — Big Data sso is the senses, Big Data is the fuel, and artificial Intelligence is the brain to realize the juture of a smart connected world.” jo? is about devices, data and connectivity. The real value of Internet of Things is about creating “navter products, delivering intelligent insights and providing new business outcomes. As millions of {jovives get connected, internet of things will trigger a massive inflow of Big Data. The key challenge is \isualising and uncovering insights from various types of data (structured, unstructured, images. contextual, dark data, real-time) and in context of your applications. I believe deriving intelligence from Big Data using artificial Intelligence technologies is the key enabler for smarter devices and a connected world. Below Diagram is showing the IOT data processed by Big Data Analytics. * Capture * 4 * Integrate - Explore Mine rs At Pico Ag ya a a \ Analytical bed fe fon Sie = ad Map Analyre # Transform * Cleanse © Share Big Data Analytics {oT Infrastructure Big Data Platform Fig. Big DataFlow inlo T ‘This section provides an overview of the parameters related to the ToT based big data and analytics ‘The categorization of these parameters is given, as like Big Data Sources, System Components, Big Data Enabling Technologies, Functional Elements, and Analytics Type. loT applications play an important role in generating big data. These ToT applications may include smart city management, Intelligent Transport System (ITS) and sensors. ‘As in smart city management system, a massive amount of data is generated through sensors, cameras, to provide security to the citizens. So definitely this data needs to be analyzed and processed to collect useful information. Similarly, smart industries use loT sensors to maximize the productivity if their products, In smart industries for the designing of new and advanced material products loT thased big data and analytics solutions have been used. The overall goal in all smart environments is to extract the valuable insights from massive volumes of data. After this extraction of valuable insights, decision makers improve the service management. Below figure Represents some Big Data Sources, System Components, Big Data Enabling Technologies, Functional Elements, and Analytics INORTH Internet of Things and Artificial intelligence parameter. of ToT haved Big Data AnulySen ‘Type. Furthermore, their classification could be dnderstood by the same. Both ToT and Big Data analytics are beneficial for each other. Io applications such as Smart Transportation, Smart Healthcare, and Smart Grids have been eereasingly adopted applications in smart environments due to the Big Data and Analytics. These spplications are taking advantage of big data and analytics as shown in Fig, 4. Taxonomy of Big Data and Analytics Solutions for loT Machine Learning (ML) Techniques e.g. Classifications, liner regression etc. In Machine Learning there are different models that generally fall into 8 different categories: (1) Supervised Learning (2) Unsupervised Learning (3) Reinforcement Learning. 1. Supervised learning: Involves an output label associated with each instance in the dataset. This output can be discretelcategorical (red, dog, panda, ford mustang, STOP sign, spam...) or real-valued. Right now, almost all learning is supervised. Your data has known labels as output. It involves & supervisor that is more knowledgeable than the neural network itself, For example, the supervisor Meads zome example data shout which he supecvisar ulready kmows the answers. The supervisor guides the system by tagging the output. For example, @ supervised machine learning system that can Iearn which emails are ‘spam’ and which are ‘not spam’. The algorithm would be first trained with available input data set (of zillions of emails) that is already tagged with this classification to help the machine learning system learn the characteristics or parameters of the ‘spam’ email and distinguish it raw those of ‘not spam’ emails. Just as a three-year-old learns the difference between & ‘block’ and @ vcoft toy” the supervised machine learning system learns which email is ‘spam’ and which is ‘not spam _— ‘Antficial Intelligence (Al), Machine Learning (ML) & Deep Lear DL) a7 Introduction 19 J) Techniques such as linear or logistic regressions and decision tree classi ‘echniques * category of learning. + Labeled data * Direct feedback + Predict outcomenuture + No labels + No feedback + “Find nidden structure* + Decision process + Reward system + Lear seres of actions Regression: This is a type of problem where we need response values, Some examples are what is the price of house in a specific city with 3 bedrooms and above 2,000 sqft? Predicting financial results, stock prices or how many total runs can be on board ina crieket game. You have an existing data set & outputs (supervised learning) and your algorithm predicts the outcome based on a fitting function. to predict and forecast for the continuous- Classification: Where you need to cates gorize a certain observation into a group. In the below picture, if you're siven a dot you need to classify it as either a blue dot or a red dot. Few more examples would ‘o predict if a given email is spam or not spam? Is a detected particle a Higgs Boson or a normal sub-atomic particle? Assigning a certain news article into a group — Will it rain today or not? Is this picture a cat or not? Detecting fraud insurance under writing be — like sports, weather, or science. or evaluating risk for frauds or Classification Regression Survived (years) Cele Genet . Unsupe known ou pursed Learning — This is an ‘unaided’ type learning when your data typically has no PUE labels op any feedback loop. This is useful when there is no example data set with known INORTH Internet of Things and Artificial Intelligence 68 Jhing for a hidden pattern. In this case, clustering ive. dividing a set of s and your are seard elements into groups according to some unknown pattern is carried out based on the existing data sets, rom the data set we provide. In general, unsupervised learning is ‘The system has to understand itself fi «bit difficult to implement and thus it’s not used as widely as supervised learning. Most popular types are clustering and association as below. answer Original unclustered data Clustered data ‘A computer-generated program showing k-means clustering Clustering: This is a type of unsupervised learning problem where we group similar things together. Some examples are: Given news articles or books, cluster them into different types of themes. Given @ set of tweets, cluster them based on cont ed for politics, health care shopping, real estate ete. tent of tweet. Could also be ust ‘Association: An association rule is where carne vogtsies you would be discovering the exact rules that oe ‘will describe the large portions of your data. “sonoma P O re esmpecet Example: People who buy X are also the ones ones Gypcassescreese who tend to buy Y. We may encounter it when a ae | we receive a book or movie recommendation 3 8 oe | based upon previous purchases or searches: suey 6. | Iso used for market — #™™-O > pene locos ‘These algorithms are al basket analysis using our © retailers shopping (point of sales) data. Shortly given many baskets, association techniques helps us understand which items inside a basket predict another item in the same basket. Associations between selected items using a data set on an actual grocery transaction over 30 days. Larger circles imply higher support, nline or offline _ ~~ ion to Artificial Intelligence (Al), Machine Learning (ML) & Deep Learning (DL) 89 introduc . ite vod crcles imply higher lif. ic. The most popular transaction was of pip and tropical fruits whi hile lv many people buy sausage along with sliced cheese, 3, Reinforcement Learning (RL)- Now instead of telling the child which toy to put in which box, you reward the child with a ‘big hug’ when the child makes the right choice or you make a ‘sad face’ when | the child makes the wrong action. Very quickly after a few iterations the child learns which toys need to go into which box — this is called Reinforcement Learning. Systems are trained by receiving virtual |,

    You might also like