ch4 green book

• Industrial Internet


  Subset of the broader horizontal classification called the Internet of Things, focusing on energy, healthcare, manufacturing, public sector, transportation, and related industrial systems

  
• Industrial systems (ISs)


  Interconnected (M2M) systems that combine sensors, actuators, logic components, and networks
• Industrial Internet systems (IISs)

  ISs that become connected to the Internet and integrate with enterprise systems, for the purpose of enhanced business process flow and analysis
• IISs


  • Provide operational data via sensors to enterprise back-end systems for advanced data processing and cloud-based advanced historical and predictive analytics
  • Advanced cloud services will drive optimized decision-making and operational efficiencies and facilitate the collaboration between autonomous industrial control systems
• Industrial Internet Reference Architecture

  Standard-based, open and widely applicable architectural framework to accelerate the widespread deployment of IISs
• Industrial Internet Reference Architecture

  • Flexible and covers a wide range of deployment scenarios across many industries
  • De-coupled from technical specifics and complexities, transcending today's available technologies
• Industrial Internet Architecture Framework (IIAF)

  Based on ISO/IEC/IEEE 42010:2011 standard, which codifies the conventions and common practices of architect design
• Stakeholder


  An individual, team, organization, or anyone who has an interest in the system
• Concern


  A topic of interest
• Viewpoint


  A way of describing and addressing the concerns
• Industrial Internet Viewpoints


  • Business
  • Usage
  • Functional
  • Implementation

  
• Business Viewpoint


  Addresses business-oriented concerns such as how the system delivers value to the business and how it aligns with business strategy, as well as financial concerns such as expected return on investment (ROI)
• Usage Viewpoint


  Takes the business requirements and realizes them through creation of user and system activities that deliver the required outcomes and business objectives
• Functional Viewpoint


  Addresses the stakeholders' concerns regarding the functionality of the Industrial Internet system
• Functional Domains


  • Control domain
  • Operations domain
  • Information domain
  • Application domain
  • Business domain
• Control Domain


  A collection of functional units that perform tasks such as reading data from sensors, applying logic and rules, and applying feedback to machines to control the process
• Control Domain


  • Requires accuracy and resolution, with logic and compute elements situated close to the sensors
  • Includes functions for sensor reading, intelligent actuation manipulation, and communication to connect sensors, actuators, controllers, and gateways
• Communication Function


  Connects all functional sensors, actuators, controllers, remote I/O devices, and gateways to a common protocol, requiring translation between legacy and new protocols
• Modeling


  Representation of the states, conditions, and behaviors of the system under control, with complexity depending on the system
• Asset Management


  Enterprise operational function that enables control systems to onboard, configure, set policy, and perform self-diagnosis and automated upgrades
• Executor


  Building block responsible for assuring policy, understanding condition and state, monitoring the environment, and providing feedback control
• Executor


  • Can be straightforward (simple feedback control) or sophisticated (incorporating cognitive and machine learning for high autonomy)
• Operations Domain


  Consists of building blocks for prognosis, optimization, monitoring and diagnosis, provisioning and deployment, and system management, which relate to and support the control domain
• Operations Domain


  • Enables remote, secure, and cost-effective provisioning and deployment of assets
  • Provides monitoring, diagnosis, and prognosis capabilities using Big Data and predictive analytics
  • Aims to improve asset performance, reliability, and efficiency through optimization
• Information Domain


  Transforms operational data into information that can be used to enhance business processes and decision-making
• Big Data lakes


  Used to determine historical and even predictive analytics, to identify potential issues with machinery and systems before they occur
• Optimization


  A set of functions aimed at improving asset performance and efficiency, to improve asset reliability and performance while reducing energy consumption, and increase availability and efficiency as well as ensuring assets operate at peak optimal performance
• Conditions required to obtain the goal of optimization

  1. Automated data collection, validation, processing, and analytics

  2. Capture and identify major system alarms, events, and issues such as power failures, system and network faults, latency, packet loss and downtime as well as non-responsive applications

  3. The intelligence and capability to analyze and determine root causes for known problems
• Information domain


  The objective is to transform the operation data harvested from the other domains, predominantly the control domain and its vast array of sensors, into information, which can be subsequently used as control feedback to stabilize or improve the process. The goal is to collect sensor data, store and process it as it can be transformed from raw data to information, and then distilled into knowledge
• Transforming raw data into knowledge greatly enhances our ability to adapt and improve operational and business decision-making, and optimizing the process functions, which delivers greater productivity, profits, and operational efficiencies
• Knowledge is always helpful as it allows managers to adapt to the present conditions, such as heavy snow that has closed normal routes, so they can reroute deliveries or be proactive in advising customers of imminent delays
• Data


  Functions for ingesting data and operational state from all the other domains; responsible for the quality-of-data processing such as data filtering, cleansing and removal of duplicate data; functions that enable syntactical and semantic transformation; provides data storage, persistence, and distribution, and this relates to batch or streaming analytics
• All these data functions can be performed in real-time streaming mode whereby the data is processed as they are received in quasi-real-time. However, this is not always possible or economical as commonly the data is stored for later batch job analysis
• Analytics


  Contains a set of functions that enable data modeling, analytics and rule engines of advanced data processing. The analytic functions can be performed on/offline in streaming or batch mode. Typically, in streaming mode, the resulting events and alerts produced are fed into the functions of the application domain, while the results of batch job analytics go to the business domain
• Application domain


  Responsible for the functions that control logic and deliver specific business functionalities. At this level, applications do not control machine or system processes, as that is the function of the control domain. Applications do not have the direct access, control, or the authority to control SSIs processes. Instead, application functions perform advisory functions; however, they do provide use-case rules and logic, as well as APIs whereby an application can expose its functionalities and capabilities to a system
• Business domain


  Enables integration and compatibility between Industrial Internet system functions and enterprise business systems such as ERP (enterprise resource management), CRM (customer relationship management), WSM (warehouse stock management), and many others
• Implementation viewpoint


  Involves technology rather than business or usage functions. IIS architecture will be guided by business and financial constraints, which will include, among other things, business requirements, alignment with corporate strategy, budget, and return on investment with regard to both financial and competitive advantage
• Architectural topology


  Follows a three-tier design, has gateway media-convertors at the border areas to convert and translate heterogeneous protocols and technologies. Furthermore, it has a multi-tier storage topology that can support a pattern of performance tier, capacity tier, and archive tier that relates to the fog and cloud, respectively. In addition, the architecture also supports edge-to-cloud (fog) direct connectivity and a distributed analytics (cloud) topology
• Edge tier


  Where data from all the endnodes is collected, aggregated, and transmitted over the proximity network to a border gateway. Depending on the protocols and technologies used within the edge tier, some data translation and interface integration may be applied at hubs, remote I/O devices, or protocol convertors. The edge tier contains the functions for the control domain
• Platform tier


  Receives data from the edge tier over the access network and is responsible for data transformation and processing. The platform tier is also responsible for managing control data flowing in the other direction, for example, from the enterprise to the edge tiers. It is within the platform tier that we will locate the majority of the functions related to the information and operations domains