Smart Cities Intelligent Traffic Management Intelligent .
white paperSmart CitiesIntelligent Traffic ManagementIntelligent TrafficManagement Edge AnalyticsWipro Intelligent Traffic Management (ITM) provides edge analytics to enablereal-time traffic monitoring. The solution utilizes OpenNESS.IntroductionMore cars are on the road than ever before, which, combined with aginginfrastructure, is making traffic conditions in urban environments considerablyworse each year. Roads and highways are plagued with congestion, causing slowmovement and long commutes. Drivers are distracted by smartphones despitenew laws that aim to limit this distracted driving. Additionally, pedestrian traffic,accidents, and traffic rule violations all add to challenging traffic conditions.Given these deteriorating traffic conditions, there is a critical need for automaticallydetecting traffic law violations, processing these violations at the edge, and enablingguaranteed electronic delivery of the traffic violation report to the traffic police inreal time.Traditionally the traffic management system is run in the cloud, where the server ishosted. For high-precision video analytics, high quality video traffic is sent to thisserver and analyzed. The round-trip delay to a central cloud server and back addsto latency, and the backhaul traffic over internet links lack service level agreements(SLAs) for predictable performance. High-throughput 5G networks are enablingedge analytics, allowing the processing of video streams locally to minimizebackhaul traffic.Intelligent traffic management (ITM) solutions are being implemented intourban transportation infrastructure and into vehicles to improve traffic flow andhelp ensure the safety of drivers and pedestrians. A key enabling technology isvideo surveillance of high traffic areas, which allows government transportationemployees to monitor road conditions so that public safety officers can betterrespond to and manage incidents and traffic flow.Table of ContentsIntroduction . . . . . . . . . . . . . . . . . . . . . 1OpenNESS Overview . . . . . . . . . . . . . 1Overview of Intelligent Traffic . . . . .Management (ITM) by Wipro . . . . . . 2End-to-End Solution . . . . . . . . . . . . . .Architecture. . . . . . . . . . . . . . . . . . . . . 3Data and Control Flows . . . . . . . . . 4Conclusion. . . . . . . . . . . . . . . . . . . . . . . 5About Wipro . . . . . . . . . . . . . . . . . . . . . 6About Intel Network Builders. . . . . 6Table of Abbreviations. . . . . . . . . . . . 6ITM systems examine this traffic video footage using video analytics to classifythe traffic based on a set of configured rules. This classification is used to detectviolation of the configured rules. The ITM solution can cross-correlate the violationwith personal identification, if permitted by law. An alert notification/output reportof defaulters is then sent to the central monitoring office, where fines are imposed.Wipro, an Intel Network Builders ecosystem member and Intel Network BuildersEdge Partner, has developed an ITM solution that utilizes an edge analyticsapplication running on the Open Network Edge Services Software (OpenNESS)platform deployed on multi-access edge computing (MEC) servers. The Intel Distribution of OpenVINO toolkit is also used to accelerate the ITM videoanalytics solution.OpenNESS OverviewOpenNESS is an open source software toolkit that enables highly optimized andperformant edge platforms to onboard and manage applications and services withcloud-like agility across any type of network.
White Paper Intelligent Traffic Management Edge AnalyticsWipro uses OpenNESS to add orchestration features toits network edge–deployed ITM software. The Wipro ITMsolution leverages the ability of OpenNESS to deploy andmanage the onboarding of traffic video analysis VNFs.OpenNESS provides application lifecycle managementfeatures that can be deployed and managed on differentedge servers located on premises or at the telco networkedge. The service orchestration capability of OpenNESSenables deployment of the solution at scale by orchestratingwith multiple edge instances of the service based on theneed. A learning model obtained from one node can bepropagated across all instances of the service.Edge NodeOpenNESS EdgeNode MicroservicesSome of the key services and features that are provided byOpenNESS in Wipro’s ITM solution include the following: Support for multiple access technologies: Works with5G, LTE, Wi-Fi, and wired networks. Edge orchestration: Exposes northbound APIs that acentral orchestrator such as ONAP can use to federateedge orchestration. Deployment: Can be implemented at the on-premisesedge or the network edge. Hardware abstraction: Supports a template for resourcedescription that simplifies deployment.Edge ControllerOpenNESS ControllerMicroservicesFigure 1. OpenNESS overview.An OpenNESS subsystem consists of one or more OpenNESSedge nodes and a controller node. Both nodes host specificmicroservices, arranging an application’s requirements asa collection of independently deployed services. An edgenode hosts a set of OpenNESS microservices, edge computeapplications, and network functions. The OpenNESS edgenode microservices deliver the following functionality: Management of application lifecycles Enforcement of DNS and network policy Steering data plane traffic to edge node applications Steering data plane traffic to local breakout (LBO) hoststhat may be attached to the edge node Supporting microservices or enhancements that exposeplatform capabilities, such as Enhanced PlatformAwareness (EPA), to the edge compute applications andnetwork functionsThe OpenNESS edge node runs on a real-time kernel andleverages the open source Data Plane Development Kit(DPDK) to accelerate the data plane implementation.The controller, depending on the deployment, eitheroperates the edge nodes by invoking the edge node APIs oruses the existing orchestrator to manage the edge nodes.The controller exposes APIs to allow network orchestratorsto operate the OpenNESS subsystem.An OpenNESS application can be categorized as followsdepending on the servicing of end user traffic: Producer application: Provides services to otherapplications running on the edge compute platform.Producer applications do not serve end user traffic directly. Consumer application: Serves end users traffic directly.Consumer applications may or may not subscribe to theservices from other producer applications on the edgenode.Wipro’s ITM services the end user traffic and therefore usesthe consumer application deployment option of OpenNESS.This producer/consumer categorization allows a servicesubscription-based application architecture.OpenNESS supports two environments for building an edgeplatform. The first one is based on a Kubernetes environment,where the controller services are run as part of theKubernetes master, while the edge node’s services run inthe cluster worker node. The second is based on a KVM andnative Docker run-time environment for the edge node, andthe controller runs in a separate node. The ITM use caseuses the latter environment to deploy and manage the edgeapplications.Overview of Intelligent Traffic Management(ITM) by WiproWipro’s edge network–based ITM solution monitors urbantraffic conditions through real-time video surveillanceanalytics. The ITM consists of the following components: A video analytics engine (VAE) using a traffic regulationpolicy database to detect traffic violations. A deep neural network model based on the low-latency,low-power, and lightweight MobileNet mobile-firstcomputer vision models, and the Intel Distribution ofOpenVINO toolkit for detecting vehicles in the videoframes. The OpenVINO toolkit is based on convolutionalneural networks (CNNs).2
White Paper Intelligent Traffic Management Edge AnalyticsIn the ITM solution, video data traffic is processed by avideo analytics virtual application running at the edge node.Wipro has benchmarked1 the efficiency of the OpenVINOtoolkit-based edge solution with increasing frame capturerates. Figure 2 depicts just how many cars can be detectedas the higher frame processing rates impact the computepower.A single street may have many different traffic parameters,including different speed limits, a combination of stoplightsand stop signs, and other traffic rule differences. Giventhe challenge of the myriad of different traffic rules andregulations for different streets, Wipro has developed atechnique of applying dynamic traffic regulations based onthe video stream received.Variation of Model Efficiency with Varying Frame RateNumber of vehicles ame capture rate (frames/second)Figure 2. Wipro’s benchmarking on model efficiency with increasing frames capture rate (higher is better). See backup forconfiguration details. For more complete information about performance and benchmark results, visit www.intel.com/benchmarks.The video analytics engine anchors the violation detectionsystem with a catalogue of traffic rules and regulations suchas speeding, running a stop sign, going in the wrong direction,and others. This engine generates a violation report after theanalytics, which is then transmitted to a central location forautomatic ticket generation.End-to-End Solution ArchitectureThe integration of OpenNESS into the ITM solution providesedge video analytics deployment as shown in Figure 3.The ITM application is packaged as a VM. This VM is thenonboarded into the OpenNESS edge node via the OpenNESScontroller. The upstream interface is connected to the videofeed. The downstream interface is connected to the PDN(packet data network).Central OfficeOpenNESS t VideoStreamViolation ReportOn-Premises EdgeViolation ReportCentral DatabaseOn-Premises EdgeOpenNESS PlatformVideoAnalyticsEngineInputVideoFigure 3. Block diagram of Wipro’s traffic video analytics solution.3
White Paper Intelligent Traffic Management Edge AnalyticsAs shown in Figure 4, the edge node has four functions usedfor ITM implementation:2. OpenNESS Enhanced Platform Awareness Microservices:These microservices include edge authentication agent(EAA), edge virtualization agent (EVA), edge lifecycle agent(ELA), syslog, DNS, and others. They manage applicationlifecycle, DNS resolution, application enrollment, and more.1. OpenNESS Data Plane Services: This steers traffictoward applications running on the edge node or thelocal breakout port. Traffic policies are configured onthe OpenNESS edge controller and pushed to the dataplane services, such that traffic steering is applied toeither redirect the traffic to edge applications for furtheranalysis or pass the packets through the downstreaminterface to the packet core for traffic forwarding over thepacket network.3. Evolved Packet Core (EPC): The edge node is attached tothe SGi interface of an EPC. Traffic from the EPC arrives asIP traffic and is steered as needed to edge applications.The EPC combines both user and control plane.4. ITM Application: Video analytics engine that runs as a VMor a Docker henticationRegistrationedgecontroller ccContainerOther OpenNESSEnhanced PlatformAwarenessMicroservicesLifecycle APIUIMySQLTraffic RoutingConfigurationForward Edge/Non-Edge TrafficUpstreamInterfaceIntel EthernetNetworkAdapterI350ContainerOpenNESS DataPlane ServicesIntel S Linux 7 .6Intel Xeon D processorSGiEPCVideoFeedUserTrafficData PathControl PathFigure 4. OpenNESS integration with ITM application.Data and Control FlowsITM integration with OpenNESS has two flows as shown inFigure 4.1. Data Flow: As per the 3GPP standards, the video feedterminates on a base station (eNodeB for 4G or a gNodeBfor 5G) and then connects to the OpenNESS edge nodevia an SGi interface (4G LTE) or an N6 interface (5G). In theITM implementation, video feeds from the traffic camerareach the ITM application on the edge node throughthe upstream interface. The video traffic is interceptedby the data plane of the edge node data plane services.On the OpenNESS edge controller, using edge nodeand application lifecycle management functionality,traffic policies are configured such that traffic steering isapplied to either redirect the traffic to edge applicationsfor further analysis or pass the packets through thedownstream interface to the packet core for trafficforwarding over the packet network.2. Control Flow: The application is authenticated with theedge node appliance (which includes ELA, EVA, EDA,and EAA) and the edge controller. Once authenticationis successful, the application is registered. In lifecyclemanagement, the ELA communicates with the edgecontroller to control the status of the application (start,stop, delete, etc.). Traffic routing configuration is definedon the controller. Based on the traffic, policy routingdecisions are made: incoming video camera traffic isredirected to the ITM and other traffic is sent directly tothe packet core without any ITM processing.4
White Paper Intelligent Traffic Management Edge AnalyticsFigure 5 shows how ITM is deployed on the edge node.Once the interfaces are declared as user plane interfaces,the ITM application VM is deployed on the edge node usingthe controller UI. Once the ITM VM status is changed torunning state, the source filter is applied on the ITM VM toredirect traffic to the ITM application. This processing is doneApplicationStep 1Nodesuccessfullyregistered andlisted oncontroller UIStep 2OpenNESSdata planeinitializedStep 3Template withapplicationhttps linkStep 4Deployed,then runningstateStep 5Policy-basedfilteringStep 6entirely at the edge and any detected traffic violations aresent to traffic monitoring personnel. This solution uses IntelDistribution of OpenVINO toolkit image processing librariesto process the video frames and detects speeding, wrongdirection driving, or other traffic violations.Application Image Creationas Docker/VMThe video analytics application ispackaged as a Docker container ora VM and hosted on the controllerusing https protocolRegistration of Edge NodeThe OpenNESS edge node server isregistered with the controller using aunique serial key generated by theedge node Ansible scriptsDeclaration of InterfacesInterfaces listed on the UI are declaredas userspace ports and given afallback interface, for upstream anddownstream trafficApplication Template CreationAn application template containingthe required hardware resources iscreated, which helps in mass scalingof the applicationDeployment/Runningof ApplicationUpstream port connected to videofeed and downstream portconnected to EPCTraffic Policy SetupOnce connection is establishedbetween the consumer and thevideo source, source/destinationfilter applied; output is visibleto the end userOutput of theProcessFigure 5. Flow diagram for application deployment on edge.ConclusionWipro has developed the ITM solution leveraging edgeanalytics to utilize traffic camera networks for real-time alertsin a way that reduces backhaul network bandwidth utilization,reduces latency, and improves scalability so that the learningmodel obtained from one node can be propagated across allinstances of the service. By using OpenNESS, the processingof the video analytics can be offloaded to edge computenodes, allowing video processing locally to minimize backhaultraffic. Orchestration can be leveraged to replicate the ITMsolution across multiple edge instances based on networksize and need.The Wipro ITM solution enables improved traffic flow and roadconditions because the ITM can be trained to detect relevantactivity and send an immediate alert to traffic monitoringpersonnel instead of transporting full video traffic to the datacenter for analysis. The ITM enables real-time notifications totraffic monitoring agents, providing new possibilities for betterroad management, critical notification during accidents, trafficviolation reporting, and more.5
White Paper Intelligent Traffic Management Edge AnalyticsAbout WiproAbout Intel Network BuildersWipro Limited (NYSE: WIT, BSE: 507685, NSE: WIPRO) isa global information technology, consulting, and businessprocess services company. It harnesses the power of cognitivecomputing, hyper-automation, robotics, cloud, analytics, andemerging technologies to help our clients adapt to the digitalworld and make them successful. A company recognizedglobally for its comprehensive portfolio of services, strongcommitment to sustainability and good corporate citizenship,it has over 160,000 dedicated employees serving clientsacross six continents. Together, its employees and clientsdiscover ideas and connect the dots to build a better and abold new future.Intel Network Builders is an ecosystem of infrastructure,software, and technology vendors coming together withcommunications service providers and end users toaccelerate the adoption of solutions based on networkfunctions virtualization (NFV) and software definednetworking (SDN) in telecommunications and data centernetworks. The Network Edge Ecosystem is a new initiativegathering ecosystem partners with a focus on acceleratingnetwork edge solutions. As an integral part of the broaderIntel Network Builders program, this initiative aims tofacilitate partners’ access to tested and optimized solutionsfor network edge and cloud environments. Learn more ystemTABLE OF ABBREVIATIONS3GPP3rd Generation Partnership ProjectITMIntelligent Traffic ManagementAPIApplication Programming InterfaceLTELong Term EvolvedCNNsConvolutional Neural NetworkMECMulti-Access Edge ComputingDPDKData Plane Development KitNFVNetwork Functions VirtualizationEAAEdge Authentication AgentONAPOpen Networking Automation PlatformEDAEdge Dataplane AgentOpenNESSOpen Network Edge Services SoftwareELAEdge Lifecycle AgentSDNSoftware Defined NetworkingEPCEvolved Packet CoreVAEVideo Analytics EngineEVAEdge Virtualization AgentNotices & Disclaimers¹ Testing conducted by Wipro in Oct. 2019: System utilized 2.10 GHz Intel Xeon Silver 4216 CPU (microcode 0x5000024) with 256 GB of DDR RAM, 889 GB hard drive and Intel Ethernet ServerAdapter I350T. The system BIOS was version 3.1, Intel Hyper-Threading Technology was turned on and Intel Turbo Boost Technology was turned off. The core OS was CentOS Linux 7 withLinux 3.10 as the kernel.Intel technologies may require enabled hardware, software or service activation.Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors.Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors maycause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that productwhen combined with other products. For more complete information visit www.intel.com/benchmarks.Performance results are based on testing as of dates shown in configurations and may not reflect all publicly available u pdates. See backup for configuration details. No product or componentcan be absolutely secure.Intel’s compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2,SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufacturedby Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are r
OpenVINO toolkit for detecting vehicles in the video frames. The OpenVINO toolkit is based on convolutional neural networks (CNNs). White Paper Intelligent Traffic Management Edge Analytics Figure 1 .OpenNESS overview. Wipro uses OpenNESS to add orchestration features to its network edge–deployed ITM software. The Wipro ITM
to technology suppliers to Smart Cities, Smart City program managers, academics researching innovation and Smart Cities, and disability organizations and advocates working to make innovation and Smart Cities more inclusive. This Playbook has been designed to complement other tools that make up the . Smart Cities for All . Toolkit.
wider cyber, digital, smart, intelligent cities texts [2]. Intelligent cities, or intelligent spaces more generally, refer to a wide range of electronic and digital applications related to digital spaces of communities and cities, for example smart grids, smart meters, and other infrastructure for electricity, water supply, and waste management.
Agenda and Target Discussion -Smart City What Smart City Components SHARED LIFE Modular Data Centers City Dashboard GIS CONVENIENT LIFE Smart Transportation - ITMS, Smart Parking, Integrated Public Transit, VMS, Traffic enforcement Smart Power - Intelligent Inspection Management System, Intelligent O&M systems .
smart grids for smart cities Strategic Options for Smart Grid Communication Networks To meet the goals of a smart city in supporting a sustainable high-quality lifestyle for citizens, a smart city needs a smart grid. To build smart cities of the future, Information and Communications Techn
Smart City Platform Platform Platform Service Application IoT World Bank Korea Week 2020 Smart Cities of Korea. 2 Trends. Gen 1 : Sustainability Development of Smart City 3 2017 Google, Sidewalk Master Plan 2014 Singapore, Smart Nation 2012 China, announced a plan to build 320 smart cities 2018 Korea, National Pilot Smart City 2011
the destination. The traffic light system designed by Salim Bin Islam provided a design and development of a microcontroller based intelligent traffic control system. He proposed a new intelligent traffic control system that is to control the traffic system through traffic signal on the basis of current traffic density.
smart cities in Cape Town, South Africa and the privacy concerns associated with smart cities. C. HARACTERISTICS OF . S. MART . C. ITIES Cities around the world are attempting to transform into smart cities, in order to be more economi-cally competitive and promote sustainable gr
2019), the term "smart city" has not been officially defined (OECD, 2019; Johnson, et al., 2019). However, several key components of smart cities have already been well-established, such as smart living, smart governance, smart citizen (people), smart mobility, smart economy, and smart infrastructure (Mohanty, et al., 2016).