Presentazione Di Powerpoint

Fog Computing, its Applications in Industrial IoT, and its Implications for the Future of 5G

Flavio Bonomi, CEO and Co-Founder, Nebbiolo Technologies IEEE 5G Summit, Honolulu, May 5th, 2017 Agenda

• Fog Computing and 5G: High Level Introduction • Architectural Angles in “Fog” with Relevance to 5G • Fog Computing and 5G: Natural Partners for the Future of Industrial IoT, with Applications • Nebbiolo Technologies: Brief Introduction • Conclusions

Fog Computing Mobile Edge Computing (Modern, Real-Time Capable) Edge Computing Real-Time Edge Cloud

1. Clouds 2. Enterprise Datacenters 3. Traditional and Embedded Public and Private Endpoints Communications 4. Networking Infrastructure (Internet, Mobile, Enterprise Ethernet, WiFi) Enterprise Data Center The Internet of Things Brings Together Information Domain and Operations Domain through: Traditional ICT end-points Information Technologies 1. Connectivity 2. Data Sharing and Analysis 3. Technology Convergence Operational Technologies

Machines, devices, sensors, actuators, things

The Future 5G networks and IoT require more virtualized, scalable, reliable, Public and Private Communications secure, real-time capable Computing Infrastructure and Storage at the Edge: (Internet, Mobile, Enterprise Ethernet, WiFi) Enterprise Fog Computing! Data Center Mobile Edge Computing! Real-time Edge Computing Traditional ICT end-points Information Technologies

Operational Technologies

Machines, devices, sensors, actuators, things

• Global IoT spending (CAGR) of 15.6% over the 2015-2020 forecast period, reaching $1.29 trillion in 2020

• Worldwide spending on the Internet of Things (IoT) is forecast to reach $737 billion in 2016

• In 2016 IoT spending led Industrial IoT verticals: Manufacturing ($178B), Intelligent Transportation ($78), and Utilities ($69B) Source: IDC Report http://www.idc.com/getdoc.jsp?containerId=prUS42209117

The spending includes: hw, services, sw, and connectivity 6 © Nebbiolo Technologies What is Fog Computing ? The missing link between Clouds and Endpoints

Fog Computing brings:

Cloud-inspired computing, IT storage, and networking Fog Computing is the key functions closer to the enabler of a real convergence data-producing sources between IT OT and OT technology while integrating real-time and safety capabilities required in the OT domain

1. Communications, gateway networking convergence Cloud 2. Edge data management, analytics 3. Distributed application hosting Fog 4. Virtualization of all resources, multi-tenancy 5. Security and Privacy 6. Real-time, local control 7. Scalability A Consortium with close to 8. Reliability 100 Members Already!

Peter Levine (Andreesen&Horowitz) during Gartner keynote session last week: “the cloud computing is dead, the intelligence/processing is IoT Endpoints going down close to the things”

Adopting Important Bandwidth, Deterministic, Low Latency, Scale and Coverage Requirements (E.g., Massive IoT, Autonomous Vehicle Communications, High Reliability,…)

Key Values of 5G: • Licensed spectrum • Reliability • Range of features • Investment • Political power

Ref: Qualcomm

IT to OT Convergence Hierarchical Data Management and Analytics Virtualized and Distributed Application Platform The Evolution of Control Decentralized Security

Information Technologies: Cloud Virtualization and Multi-Tenancy, Software Management Automation, Fog IT Data Analytics, Scalability, Software Defined Networking (SDN), Security and Privacy OT

Operations Technologies:

Real-time, Safety, Reliability, Control, Machine Connectivity and Data Acquisition, Human Machine IoT Endpoints Interface

Information Networking Technologies: Cloud Ethernet, WiFi, Cellular 3/4/5G, Bluetooth LE, SDN Fog

Operations Networking Technologies:

Real-time capable and Safety capable Field Networking Industrial Wireless, IEEE TSN (Deterministic Ethernet), LoRA, PLC

IoT Endpoints

Wireless Technologies: Wired Technologies:

Distance 1. Cellular (2G/3G/4G/5G) IoT and Wireless Technologies 1. Deterministic Ethernet 10,000m Map 2. Bluetooth Low Energy (BLE) TIME-SYNCHRONIZED( LoRa

REAL(TIME( 3. LoRa (Low power, long range, low MANAGED( bandwidth) 1000m

UNMANAGED(( HSPA ( 4. IEEE 802.15.4 with 6TSCH 5G LT 100m 802.11a/g E 2. Power Line Communications 5. WiFi: Low Power, Deterministic, 802.15.4 Advanced Vehicular (802.11.p) WiFi LTE 10m 802.11ac BluethoothLE Data Rate (bps) 100k 1M 10M 100M 1G 10G

Deterministic, Time-triggered Ethernet is based on:

1) A global notion of time 2) A communication global schedule (when to do what)

Network Virtualization based on Fundamental for Industrial IoT! Deterministic Ethernet Synchronous Traffic • real-time control • safety Deterministic = Reliable, Very Low Jitter, even more than Low Streaming Many Virtual Links • audio on a Single Cable Latency • video

Standard Ethernet Traffic Future Standard for • diagnostics over IP • download and updates Automotive, Transportation,

Industrial,… • Guaranteed timing for hundreds of real-time control functions • Guaranteed bandwidth even in case of excessive network loads • Guaranteed delivery even in case of network faults www.tttech.com

15 © Nebbiolo Technologies Page 14 Real-Time Fog Computing Requires Deterministic Ethernet Many Deterministic Communications Scenarios Across Distributed Real-time Applications

Between Apps within fogNodes 3 fogNode Can 5G Play Here? 2 Between Apps on a Backplane Interconnect FogLet

A1 A1 A1 A2 A1 A2 4 Intel Multicore Intel Multicore Intel Multicore Between Apps across Application 1 Application Application fogNodes Processor Processor Processor Between fogNodes and Endpoints (Fieldbus role) …. FPGA FPGA FPGA A2 with DE with DE with DE Intel Multicore Application Switching and Switching and Switching and Processor ARM CPUs ARM CPUs ARM CPUs

FPGA with Switching and ARM CPUs …. …. …. Deterministic Ethernet DE

A Switch Real-time App A2

• Zero-Touch deployment of Fog nodes, and assets

Cloud System Management • Application hosting and full Life Cycle Management

Federation • Asset Management of Fog Nodes • Management and scheduling of real-time resources

Machines • End-to-end security management and Things • Fog node federation, distributed storage

• Learning in the Cloud and Learning to simplified Cloud deploying models simpler models in the Fog

Fog

• Fast reaction to rich local analysis

100s μsecs to mins

100s μsecs to msecs • Data objectization and reduction

μ μsecs IoT Endpoints

Virtualization:

A combination of physical separation (multicore), hard, RT-NRT Virtual Board/Machine based virtualization and more lightweight Linux/Windows Container or Docker based virtualization

NRT-VM 2 RT-VM 1 RT-VM 2

AA A A

Container

Docker Docker

Win

N 1 … Win RT-Apps RT-Apps Apps Linux Windows … RT-OS 1 RT-OS 2

GuestOS+Hypervisor (KVM with RT Patches) or Thin Hypervisor (Lynx,..)

Deterministic Networking and Real-time Virtualized Computing enable the Convergence of Multiple Control Functions, one step removed from the controlled Endpoints: Compu: ng Virtualiza: on with Real-: me and Non Real-: me OSs Fog + Network Virtualization based on The Software Defined Machines! (Ref: GE) Deterministic Ethernet Determinis: c Fog Networking Synchronous Traffic • real-time control • safety Deterministic Real-time Streaming Many VirtualNetworking Links Virtualized • audio on a Single Cable • video Computing

Standard Ethernet Traffic • diagnostics over IP • download and updates

IoT Endpoints • Guaranteed timing for hundreds of real-time control functions • Guaranteed bandwidth even in case of excessive network loads • Guaranteed delivery even in case of network faults www.tttech.com

15 © Nebbiolo Technologies 13 IoT Endpoints

Fog Resources Can Be Used to Cloud Help Less Capable Devices, i.e.,: • Identify • Protect Fog • Isolate • Verify • Upgrade A Fundamental Mediation Role

DecentralizedCentralizedDistributed

- Industrial Automation - Automotive and Intelligent Transportation - Smart Grid

Motivations: Licensed spectrum, reliability, range of features, investment, ….

Hierarchical, Virtuous Cycles of Data Acquisition, Analysis and Control

Cloud level optimization

Local, orchestrated Fog Computing: Communications,r Analysis, Control and control based on Application Hosting and Orchestration rich local analysis

Fast, Light Endpoint Control

Today’s Functionality

ERP/MRP System

Manufacturing Execution Data Center/Clouds System

Cell Controller

Machine Vision and Bar Code Reader 2) Converging and Enriching Programmable Logic Current Functionality Controller (PLC) System

Robotics Controller Fog Layer And Visualization ? 1) Adding New Functionality

Cell 1 Cell N

1) Adding New Functionality Cloud

Fog Fog Layer

2) Converging and Enriching Current Control Functionality

Key Directions: Deterministic Ethernet Network Internal Networking Convergence and Consolidated, Virtualized ComputingLow Speed Virtualization Network ECU “Data Center” LIN, Lo-speed CAN Security Entertainment network Mobility Xand-by Multi-Wire/-mode MOST, internal WiFi SafetyCommunications Network Flexray (5G)

Centralization!!!Wireless in-car network, Bluetooth, Low Power WiFi, RFID Fog Node on Wheels! Collision Radar

High-speed network Hi-speed CAN Electronic Control Unit (ECU) Central Gateway Wheel-Sensor

Public Clouds Private Clouds

Virtualized, scalable, reliable, secure, real- time capable Computing and Storage at Public and Private the Edge: Communications Fog Computing! Infrastructure (Internet, Enterprise Ethernet, WiFi) Large Potential Role for 5G!

3. Platform Overview

3.1 Platform Vision

Many of today’s solutions for managing and controlling the electric power system are designed and packaged to perform a single function. They struggle to meet near-real-time speed requirements, necessitate extensive back-end integration effort, and often lack scalability. The electric power system of the future must employ multi-functional, standards-based, and modular systems to promote and allow interoperability, lower costs, and improved reliability by integrating and analysing multiple information sources based on their timeliness, location, and availability.

These new requirements demand an electric grid with “distributed intelligence” that has the potential to significantly increase the operational efficiencies of the electric power system resulting in benefits realization through additional cost savings by:

Deferring capital infrastructure expansion Achieving improved operational performance Improving system response times More effectively managing the scalability associated with field devices The Near Future of TheDriving Smartgreater insigh tElectrical for more optimal de cGridision ma kRequiresing Fog Computing and 5G! Distributed Intelligence Platform (DIP) Reference Architecture: Volume 1 DistributedPage 11 of 31 Intelligence, S tr ITeam andlining the OT statu Convergence,s monitoring and security Standard,of all communicati ngInteroperable, field assets Secure Enabling workforce management to efficiently prioritize resources Distributed Intelligence Platform (DIP) Reference Architecture: Volume 1 Page 11 of 31

and 5G

Figure 3-2: Several examples of packaging scenarios identified for reference architecture

A subset of packaging will also include options for chassis design to enable plug-and-play modular functionality for the middle to upper-tier family of nodes. A full description of the concept of tiers and their definition is included in Section 3.2.2. It is envisioned that examples of plug-and- play modFiulguesre ma 3-2y : incSeludverae: lm exodamemplses, I/O of cpaardsck,ag progring samcenmarabilose l ogideicnt cifoniedtr olforlers ref, erencpowere s aurppchliiestec, ture and expandable computing hardware. Figure 3-3 is a representation of the hardware described in proceeding sections. A subset of packaging will also include options for chassis design to enable plug-and-play modular functionality for the middle to upper-tier family of nodes. A full description of the concept of tiers and their definition is included in Section 3.2.2. It is envisioned that examples of plug-and- play modules may include: modems, I/O cards, programmable logic controllers, power supplies, and expandable computing hardware. Figure 3-3 is a representation of the hardware described in proceeding sections.

Figure 3-3: Modular Chassis Design

The package requirements for this reference architecture are defined to meet industry standards and Duke Energy’s environmental, electrical, safety, and security requirements and support the best practices for hardware installations including the aesthetics, indicators (visual and audio), dimensions (size and weight), and color.

3.1.2 Deployment Considerations Figure 3-1: Duke Energy Distributed Intelligence Platform Node Architecture The deployment considerations of theFigu norede 3 -i3s : infMoluedulncared Chas by thsei spa Desckagigning type and operational environment that it is to be installed in, ranging from an environmentally controlled data center to outdTooher padisctrkibuagtieo rneq poluierem-topsen. ts for this reference architecture are defined to meet industry standards ____an__d _D__uk_e__ E__ne_r__gy_’s__ e_nv___iro__nm__en___tal_,__ e__lec_tr__ic_a__l, s_af___et__y, __an_d__ s_ec__uri__ty_ __re_qu__i__r_em_____en__ts____ an___d_ _s___up po_rt__ the______REV 1.1 Copyright © 2015 Duke Energy Corporation All rights reserved. best practices for ©ha rNebbiolodware installati Technologiesons including the aesthetics, indicREatorsV (v i1.sua1l an d audio),Co pyright © 2015 Duke Energy Corporation All rights reserved. 33 dimensions (size and weight), and color.

3.1.2 Deployment Considerations The deployment considerations of the node is influenced by the packaging type and operational environment that it is to be installed in, ranging from an environmentally controlled data center to outdoor distribution pole-tops. ______REV 1.1 Copyright © 2015 Duke Energy Corporation All rights reserved.

The Near Future of The Smart Grid Requires Fog Computing and 5G!! A Standardization Across USA Utilities, with Proof of Concept Deployment, has been Achieved in One Year !!!

RECOMMENDATION Approved by the RMQ Executive Committee via Notational Ballot on February 4, 2016 For Quadrant: Retail Markets Quadrant

Requesters: RMQ OpenFMB Task Force Request No.: 2015 Retail Annual Plan Item 9.a/R14008 Request Title: Develop model business practices to support OpenFMB architecture for interoperable data exchange between distributed power systems devices on the electric grid’s field area networks.

1. RECOMMENDED ACTION: EFFECT OF EC VOTE TO ACCEPT RECOMMENDED ACTION: X Accept as requested X Change to Existing Practice © Nebbiolo Technologies 34 Accept as modified below Status Quo Decline

2. TYPE OF DEVELOPMENT/MAINTENANCE Per Request: Per Recommendation: X Initiation X Initiation Modification Modification Interpretation Interpretation Withdrawal Withdrawal

X Principle X Principle X Definition X Definition X Business Practice Standard X Business Practice Standard Document Document Data Element Data Element Code Value Code Value X12 Implementation Guide X12 Implementation Guide Business Process Documentation Business Process Documentation

3. RECOMMENDATION

SUMMARY: In response to Duke Energy’s request for an open standard for utility field device interoperability (R14008), and in collaboration with Smart Grid Interoperability Panel (SGIP), the NAESB OpenFMB Task Force was formed to develop such a standard. OpenFMB is a framework specification for power systems field devices to leverage a non- proprietary and standards-based reference architecture platform, which consists of internet protocol (IP) networking, internet standard messaging protocols, and standardized common semantic models, to enable the secure, reliable, and scalable communications and peer-to-peer information exchange between devices on the electric grid.

This document contains model business practices that drive the establishment of the framework specification and approaches for how OpenFMB devices can be implemented to drive field device interoperability. This document also references a set of design artifacts (XML Schemas) that provide a starting point for any party to implement and Page 1 Brief Introduction to Nebbiolo Technologies

Producing wonderful wines: Barolo, Barbaresco, Nebbiolo, Valtellina Reds

Nebbiolo Technologies is architecting and building an innovative Fog Computing Platform for IoT Solutions

and applying it, first, in the vertical of Industrial Automation

Team: World-class, Cisco sourced, experienced (20+ people) team surrounded by a rich ecosystem of IoT technology partners Investors:KUKA Robotics, TTTech and GiTV(Tokyo, Japan VC) Milestones: 7 Patents pending; Strong Traction; Production deployments and PoCs ongoing; First product released (December 2016)

1. A flexible hardware architecture manifesting in a family of fogNodes

2. A rich software distributed stack (the fogOS), enabling fast, secure, flexible communications, data management and application deployment

Fog 3. An end-to-end system management of distributed networking and System Management computing systems, assets, software and applications (the fogSM)

Business Application IoT Infrastructure Application hosting & Orchestration Federation of Midddleware fogNodes

Secure Stack Cloud Infrastructure Fog Infrastructure Admin Plane Machines

Manageability RTOS/Kernel and Host OS/Hypervisor Things

Secureboot Hardware ( X86/Arm)

Fog Computing and 5G are Natural Partners for the Future of Industrial IoT

More Reliable, Lower Latency, Deterministic Wireless Networking is Essential for Real-time Fog Computing and its Industrial Applications!

More Collaboration nadExperimentationis Required!

Let usMove Boldly, Together: The Future is Bright!

Nebbiolo Technologies, Inc. 860 Hillview Court, Suite 310 Milpitas, CA 95035 P +1 (408) 770-2828