Internet Of Things Reference Architecture & Case Studies

Transcription

Saturn 2016Internet of ThingsReference Architecture& Case StudiesYulian SlobodyanSerge Haziyev

“Look deep into nature,and thenyou will understandeverything better.” Albert Einstein

CognitiveLoopLearnSenseActOperate

enseSenseActOperateOperate

atorsHMIs

Reference ModelPeople& ThingsEdge& CommunicationDataApplications& ServicesActLinkContextualizeAugment IntelligenceSenseBridgeIngestAugment BehaviorOperateInterconnectProcessStoreManage & Control

Implementation ViewPeople& ThingsEdge& eSensorsWSNEdgeGatewaysHC IEnvironmentHC rvicesM2MBatchManage& guration SoftwareIdentity & SecurityMGNTMGNTAccess MGNT MGNTApplications& TAlerting

Quality AttributesPeople& ThingsEdge& CommunicationDataApplications& lityDurability & ncePower nteroperability

Known VulnerabilitiesThat Ship With New DevicesCleartext Local APIUnencrypted StorageBackdoor AccountsCleartext Cloud APIRemote Shell AccessUART Access

People & Things: TacticsPeople& ThingsSecurityMaintainabilityReliabilityHardware Root of Trust(TPM/SE/Kerkey/STSAFE-A)Firmware verificationDurability & ResiliencePower efficiencyInvisibilityPhysical tamperingdetectionTrusted ExecutionEnvironment (TEE)Secure onboardingand provisioningprocess (HSM)Strong encryptionStrong authenticationof interactions (network,peripherals)

Case Study:Smart Fridges SolutionDesigned to scale up to hundredsthousands devicesFreezers and fridges monitoring and management solution.Cloud based services allowing bidirectional communication to hundredsthousands sensors/actuators for data collection/remote assetsmanagement and analytics.Business ValueOperational expensesoptimizationMarketing insights based onuser behavior analyticsImproved supply chainand logistics

Cloud ArchitectureMaster DataManagementDevice ppingLWM2MUI / APIUI / APIUIDeviceConfigurationManagementImport /ExportImport /ExportMDM CoreDevice /Master DataMappingsMappingMaster DataDictionaries& ationServersStreaming &Enrichment(Spark)SpeedLayerRaw )Report(Redsift)Journal(Graphite / InfluxDB)Data collection,Reporting Operational / )

Sensors HardwareWAN: GSMGateway nodes: ARM8Leaf nodes: ARM Cortex M4Sensor Network: I2 C, RFSensors: Light, temperature, AC/DC,camera, weight, other

AnalyticsDashboard

DEMOcraticIoT

UserIcons created by Simon Child from Noun ontractOwnerSound LevelSensorsSensorsOwnerTechSupport

SummaryBack to the nature: IoT systems resemble Human as an organism andHuman society as an organization of services and devicesSEI Architectural methods such as ATAM and ADD are well applicableto IoT domain - Quality Attributes and Tactics help design andevaluate complex IoT solutionsSecurity is a significant challenge in IoT adoption, but most vendorshave started realizing it and we will see improvements soon

Questions

This session uses real-world case studies to share a vision of the current state of standardization for the Internet of Things and describes several reference architectures. Created Date 20160502211936Z