Title Page Copyright and Credits Hands-On Industrial Internet of Things Dedication About Packt Why subscribe? Packt.com Contributors About the authors About the reviewer Packt is searching for authors like you Preface Who this book is for What this book covers To get the most out of this book Download the example code files Download the color images Conventions used Get in touch Reviews Introduction to Industrial IoT Technical requirements IoT background History and definition IoT enabling factors IoT use cases IoT key technologies What is the I-IoT? Use cases of the I-IoT  IoT and I-IoT – similarities and differences IoT analytics and AI Industry environments and scenarios covered by I-IoT Summary Questions Further reading Understanding the Industrial Process and Devices Technical requirements The industrial process Automation in the industrial process Control and measurement systems Types of industrial processes Continuous processes Batch processes Semi-continuous processes Discrete processes The CIM pyramid CIM pyramid architecture – devices and networks Level 1 – sensors, transducers, and actuators Level 2 – RTU, embedded controllers, CNCs, PLCs, and DCSes Level 3 – SCADA, Historian Level 4 – MES Level 5 – ERP CIM networks The I-IoT data flow The Industrial IoT data flow in a factory The edge device The Industrial IoT data flow in the cloud Summary Questions Further reading Industrial Data Flow and Devices Technical requirements The I-IoT data flow in the factory Measurements and the actuator chain Sensors The converters Digital to analogical Analog to digital Actuators Controllers Microcontrollers Embedded microcontrollers Microcontrollers with external memory DSPs PLCs Processor module Input and output (I/O) module Remote I/O module Network module Other modules DCS Industrial protocols Automation networks The fieldbus Supervisory control and data acquisition (SCADA) Historian ERP and MES  The asset model ISA-95 equipment entities SA-88 extensions Summary Questions Further reading Implementing the Industrial IoT Data Flow Discovering OPC OPC Classic The data model and retrieving data in OPC Classic OPC UA The OPC UA information model OPC UA sessions The OPC UA security model The OPC UA data exchange OPC UA notifications Understanding the I-IoT edge Features of the edge The edge gateway The edge tools The edge computing The IoT edge versus the I-IoT edge The fog versus the I-IoT edge The edge architecture The edge gateway The edge computing The edge tools Edge implementations Azure IoT Edge Greengrass Android IoT Node-RED Docker edge Intel IoT Gateway Edge internet protocols Implementing the I-IoT data flow I-IoT data sources and data gathering PLC Advantages of the PLC Disadvantages of the PLC DCS SCADA Advantages of SCADA systems Disadvantages of SCADA systems Historians Advantages of Historians Disadvantages of Historians Edge deployment and data flow scenarios Edge on fieldbus setup Strengths of the edge on fieldbus setup Weaknesses of the fieldbus setup Edge on OPC DCOM Strengths of the edge in OPC DCOM Weaknesses of the edge in OPC DCOM Edge on OPC Proxy Strengths of the edge on OPC Proxy Weaknesses of the edge on OPC Proxy Edge on OPC UA Strengths of the edge on the OPC UA Weaknesses of the edge on OPC UA OPC UA on the controller Summary Questions Further reading Applying Cybersecurity What is a DiD strategy? People Technology Operating modes and procedures The DiD in Industrial Control System (ICS) environment Firewalls Common control-network-segregation architectures Network separation with a single firewall A firewall with a DMZ A paired firewall with a DMZ A firewall with DMZ and VLAN Securing the I-IoT data flow Securing the edge on fieldbus Securing the edge on OPC DCOM Securing the edge on OPC Proxy Securing the edge on OPC UA Securing OPC UA on a controller Summary Questions Further reading Performing an Exercise Based on Industrial Protocols and Standards Technical requirements The OPC UA Simulation Server OPC UA Node.js Starting an OPC UA sample server Prosys OPC UA Simulator Installing the Prosys server Simulating measures The edge Node-RED Summary Questions Further reading Developing Industrial IoT and Architecture Technical requirements Introduction to the I-IoT platform and architectures OSGi, microservice, containers, and serverless computing Docker The standard I-IoT flow Understanding the time-series technologies OSIsoft PI Proficy Historian Uniformance Process History Database (PHD) KairosDB Riak TS (RTS) Netflix Atlas InfluxDB Elasticsearch Cloud-based TSDBs OpenTSDB Asset registry Data-processing and the analytics platform EMAs Advanced analytics Big data analytics Cold path and hot path Summary Questions Further reading Implementing a Custom Industrial IoT Platform Technical requirements An open source platform in practice Data gateway Mosquitto as MQTT connector Apache Kafka as a data dispatcher Kafka Streams as a Rule Engine Storing time-series data on Apache Cassandra Apache Cassandra KairosDB Installing Apache Cassandra Installing KairosDB Installing the Kafka KairosDB plugin Graphite Developing our batch analytics with Airflow Installing Airflow Developing a KairosDB operator Implementing our analytics Other open source technologies for analytics Building an asset registry to store asset information Building an asset model with Neo4j Pro and cons of the proposed platform Other technologies RabbitMQ Redis Elasticsearch and Kibana Grafana Kaa IoT Eclipse IoT Other I-IoT data beyond the time-series Apache HDFS and Hadoop Apache Presto Apache Spark Summary Questions Further reading Understanding Industrial OEM Platforms Technical requirements I-IoT OEM platforms Why do we use I-IoT commercial platforms? The Predix Platform Registering to the Predix Platform Installing prerequisites Configuring the user authentication and authorization services Configuring the time-series database Configuring security Ingesting our first bit of data Getting our data Deploying our first application Predix Machine Configuring the Predix developer kit Predix Edge OS Predix Asset The other Predix services The MindSphere platform Registering to MindSphere Working with MindSphere Other platforms Summary Questions Further reading Implementing a Cloud Industrial IoT Solution with AWS Technical requirements AWS architecture AWS IoT Registering for AWS Installing the AWS client IoT Core Setting the policies Registering a thing Working with an MQTT client Storing data DynamoDB Using acts in IoT Core AWS Kinesis AWS analytics Lambda analytics Greengrass Working with Greengrass Step 1 – building Greengrass edge Step 2 – configuring Greengrass Step 3 – building the OPC UA Connector Step 4 – deploying the OPC UA Connector AWS ML, SageMaker, and Athena IoT Analytics Building a channel Building the pipeline and the data store Preparing the dataset QuickSight Summary Questions Further reading Implementing a Cloud Industrial IoT Solution with Google Cloud Technical requirements Google Cloud IoT Starting with Google Cloud Installing the GCP SDK Starting with IoT Core Building the device registry Registering a new device Sending data through MQTT Bigtable Cloud Functions Running the example GCP for analytics GCP functions for analytics Dataflow BigQuery Google Cloud Storage Summary Questions Further reading Performing a Practical Industrial IoT Solution with Azure Technical requirements Azure IoT Registering for Azure IoT IoT Hub Registering a new device Sending data through MQTT Setting up Data Lake Storage Azure analytics Stream Analytics Testing Stream Analytics Advanced Stream Analytics Data Lake Analytics Custom formatter and reducer with Python, R, and C# Scheduling the job ML Analytics Building visualizations with Power BI Time Series Insights (TSI) Connecting a device with IoT Edge Azure IoT Edge applied to the industrial sector Building Azure IoT Edge with OPC UA support Comparing the platforms Summary Questions Further reading Understanding Diagnostics, Maintenance, and Predictive Analytics Technical requirements Jupyter I-IoT analytics Use cases The different classes of analytics Descriptive analytics KPI monitoring and health monitoring Condition monitoring Anomaly detection Diagnostic analytics Predictive analytics Prognostic analytics Prescriptive analytics CBM Production optimization analytics I-IoT analytics technologies Rule-based Model-based Physics-based Data-driven Building I-IoT analytics Step 0 – problem statement Step 1 – dataset acquisition Step 2 – exploratory data analysis (EDA) Step 3 – building the model Data-driven versus physics-based model Step 4 – packaging and deploying Step 5 – monitoring Understanding the role of the infrastructure Deploying analytics Streaming versus batch analytics Condition-based analytics Interactive analytics Analytics on the cloud Analytics on the edge Greengrass and FreeRTOS Azure functions on the edge Analytics on the controller Advanced analytics Open System Architecture (OSA) Analytics in practice Anomaly detection Steps 0 and 1 – problem statement and the dataset Problem statement Preparing the environment Step 2 – EDA Step 3 – building the model Extracting the features Selecting features Defining the training set against the validation set Building the algorithm Step 4 – packaging and deploying Step 5 – monitoring Anomaly detection with ML Step 3 – building the model Predictive production Steps 0 and 1 – problem statement and dataset Step 2 – EDA Step 3 – building the model Steps 4 and 5 – packaging, deploying, and monitoring Summary Questions Further reading Implementing a Digital Twin – Advanced Analytics Technical requirements Advanced analytics and digital twins Data-driven and physics-based approaches Advanced technologies ML Supervised learning Unsupervised learning Reinforcement learning (RL) DL TensorFlow Advanced analytics in practice Evaluating the RUL of 100 engines Steps 0 and 1 – problem statement and dataset Problem statement Preparing the environment Step 2 – exploratory data analysis (EDA) Step 3 – building the model Extracting the features Selecting variables Identifying the training set and the validation set Defining the model Step 4 – packaging and deploying Step 5 – monitoring Monitoring a wind turbine Steps 0, 1, and 2 – problem statement, dataset, and exploratory data analysis Step 3 – building the model Steps 4 and 5 – packaging and deploying, monitoring Platforms for digital twins AWS Predix GCP Other platforms Advanced modeling Other kinds of I-IoT data Summary Questions Further reading Deploying Analytics on an IoT Platform Technical requirements Working with the Azure ML service Starting with the Azure ML service Developing wind turbine digital twins with Azure ML Developing the model Building the image of the model Registering the model Deploying the model Testing the model Cleaning up the resources Understanding the ML capabilities of the Azure ML service Building the surrogate model with logistic regression and Scikit-Learn Building the training model Preparing the cluster to deploy the training model Submitting the model to the cluster IoT Hub integration Implementing analytics on AWS SageMaker Evaluating the remaining useful life (RUL) of an engine with SageMaker Downloading a dataset on S3 Starting the notebook Working with the dataset Understanding the implementation of a SageMaker container Building the container Training the model locally Testing the model locally Publishing the image on AWS cloud Training the model in AWS SageMaker Testing the model on AWS SageMaker notebook Understanding the advanced features of SageMaker Consuming the model from AWS IoT Core Understanding the advanced analytics capabilities of GCP  ML Engine Discovering multi-cloud solutions PyTorch Chainer MXNet Apache Spark Summary Questions Further reading Assessment Chapter 1: Introduction to Industrial IoT Chapter 2: Understanding the Industrial Process and Devices Chapter 3: Industrial Data Flow and Devices Chapter 4: Implementing the Industrial IoT Data Flow Chapter 5: Applying Cybersecurity Chapter 6: Performing an Exercise Based on Industrial Protocols and Standards Chapter 7: Developing Industrial IoT and Architecture Chapter 8: Implementing a Custom Industrial IoT Platform Chapter 9: Understanding Industrial OEM Platforms Chapter 10: Implementing a Cloud Industrial IoT Solution with AWS Chapter 11: Implementing a Cloud Industrial IoT Solution with Google Cloud Chapter 12: Performing a Practical Industrial IoT Solution with Azure Chapter 13: Understanding Diagnostics, Maintenance, and Predictive Analytics Chapter 14: Implementing a Digital Twin - Advanced Analytics Chapter 15: Deploying Analytics on an IoT Platform Other Books You May Enjoy Leave a review - let other readers know what you think