Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Real-Time Analytics for Industries

1,322 views

Published on

Government, telecommunications, healthcare, energy and utilities, finance, insurance and automotive all have different challenges and requirements. However, all industries are facing unlimited potential to harvest all data, all the time. Stream Computing analyzes data in motion for immediate and accurate decision making

Published in: Data & Analytics
  • Be the first to comment

Real-Time Analytics for Industries

  1. 1. © 2014 IBM Corporation Real-Time Analytics for Industries: Real-Time Analytics on Data in Motion Analyze More, Speed Actions, Store Less
  2. 2. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 2 • Core components of Stream computing • For Government - Example architecture - cyber security defense • For Healthcare - Example architecture - critical care hospital • For Finance - Example architecture - client identifying data (CID) usage • For Automotive- Example architecture – real-time diagnostics • For Telecommunications -Telecommunications Event Data Analytics (TEDA) • architecture – real-time campaigns • For Insurance - Example architecture – big data telematics • For Energy and Utilities - Example architecture – smart grid • Learn More Agenda Government, telecommunications, healthcare, energy and utilities, finance, insurance and automotive all have different challenges and requirements. However, all industries are facing unlimited potential to harvest all data, all the time. Stream Computing analyzes data in motion for immediate and accurate decision making
  3. 3. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 3 Three core components of InfoSphere Streams Integrated Development Environment Scale-Out Runtime Analytic Toolkits Cloud and on premise available for flexible deployment Agile and Manageable Functional and OptimizedFlexible and Scalable
  4. 4. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 4 Continuous and speedy analysis in context for government Smarter surveillance: Analyze data from manned and unmanned vehicles and cameras to alert law enforcement of potential issues. Identification of fraud and terrorist activity: Understand identities in real time to alert officials of persons of interest. Cyber attack discovery, prediction and prevention: Analyze real-time events across multiple layers of the network traffic to find malware and track behavior. Street crime awareness: Mine data on geospatial parameters to monitor street gangs and proactively prevent crimes. Government City of Davao better anticipates impending problems and increases situational awareness about city events Protect against threats in real time and reduce fraud
  5. 5. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 5 Example architecture - cyber security defense Cyber Security Expert Trained Models Raw Data & Model Analysis Results Botnet Reconstruction, Event Correlation, Patterns OPTIONAL Base Models Network Behavior Modeling Fast Fluxing DNS Amplification Attacks DNS Poisoning DNS Tunneling QRadar SIEM Dashboard Analytic Server Server SPSS SPSS InfoSphere Streams Ingestion Enrichment Extraction Real-Time Scoring SPSS SPSS C&DS SPSS Modeler Reporting Predictive Analytics InfoSphere BigInsights Master Data Store Reports Real-Time Detection Model Analysis Training Data & Stored Models PureData for Analytics Data Warehouse Streaming Data Net Flow Data DNS logs White listed & blacklisted Geo-IP, ASN Databases PCAP Data
  6. 6. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 6 Continuous and speedy analysis in context for healthcare Identification of life-threaten conditions: Fuse different data sources in real time. Analyze physiological streams and electronic health record to spot life-threatening conditions. Highly personalized care: Detect signs earlier to improve patient outcomes and reduce length of stays. Automated or clinician-driven knowledge discovery to indentify new relationships between data stream events and medical conditions. Proactive treatment: Build a profile for each patient based on personalized data streams and receive insights in real time to improve care. Healthcare Emory analyzes 100,000 real-time data points per second Anticipate disease onset and deliver real- time patient data to make life saving decisions
  7. 7. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less Example architecture - critical care hospital Adaptation Layer O2 Saturation Trending Analysis HRV for Situational Awareness Mortality Risk Assessment Analyze InfoSphere Streams Analytical Layer Delivery Layer real-time/replay SQL/NoSQL/HTTP Intracranial Pressure (ICP) monitoring Sepsis, AFIB, seizure detection Acquire Act 5 Custom analytics and more
  8. 8. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 8 Continuous and speedy analysis in context for finance Faster trades: Automate trades in milliseconds to increase revenue. Industry knowledge: Connect to widely used market data sources and industry systems such as FIX and QuantLib to lower IT costs. Analytic accelerators: Calculate equity option derivative values to increase revenue. Real-time support: Ingest and manage data to support equities, derivates, commodity and forex trading. Incorporate additional contextual awareness (news, weather etc) into trading decisions. Manage risk in real time: Continuously monitor. Finance Financial institution picks IBM over Storm due to better performance; real-time analysis, with latency as low as 100 microseconds Lower risk, cost and fraud while enabling faster more informed transactions and greater revenue
  9. 9. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 9 Example architecture - client identifying data (CID) usage Employee access to CID intercepted IBM CID Big Data Solution Application hosting CID informationAcquire Identify CID on user’s screens Analyze information as it arrives from thousands of sources Act Trigger action in real-time based on anomaly detection Investigation Officer Operational Risk Manager RealTime Define and deploy patterns Investigation (Real-time and Retrospective) Not covered with traditional CID systems
  10. 10. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 10 Continuous and speedy analysis in context for automotive More profitable aftermarket for services and products: Create targeted offers based on driving preferences such as sound systems and entertainment to increase revenue. More interactive and safer driving experience: Alert approaching drivers of slick conditions that caused previous drivers to use anti-lock brakes. Integrated vehicle data: Share data across third parties such as insurance companies and emergency medical services to increase collaboration and lower costs. Improved quality and functionality: Detect problems sooner, predict breakdowns, and ensure parts are in stock to keep clients satisfied. Automotive Optimize operations, improve the driving experience, and create safer roadways Peugeot integrates data from cars, logs and social media
  11. 11. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 11 Example architecture – real-time diagnostics On board diagnostics Message Sight InfoSphere Streams Smartphone InfoSphere BigInsights Portal, Aggregation, Consolidation Real-time monitoring MQTT REST Calls Developers Advanced Analytics (spatiotemporal, correlations, predictive) Analysts MQTT MQTT
  12. 12. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 12 Continuous and speedy analysis in context for telecommunications Processing of call data in real time: Process CDRs and filter SMS spam in real-time to predict customer churn and fraud. Timely marketing promotions and ability to analyze success in real-time: Trigger promotions. Determine the success of promotions within minutes and take necessary corrective actions. High utilization of expensive network assets: Understand geospatial location of the callers to target them effectively. Incremental revenue from newer marketing promotions: Run powerful geospatial analytics to cross sell additional services. Telco Increase customer satisfaction, maximize asset utilization and proactively retain profitable customers Asian telco improves marketing effectiveness 600% while lowering development cost by 95%
  13. 13. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 13 IBM Telecommunications Event Data Analytics (TEDA)  Easy to use application framework and tools to ingest, transform, and enrich data records for downstream applications and systems  Ability to customize business logic  Special adapters and toolkits ready to be used with data from telco networks  Speed custom implementations and reduce development and test costs Input Data Files NFS, GPFS, FTP * Released via GitHub CDRs Data Records Lookup & Enrichment Data Metadata Checkpoints Setup Wizard Cheat Sheets Operator GUI Output Data Files NFS, GPFS, HDFS Parsers: ASN.1, binary, CSV File/Directory Operators Priority Handling Queue (S)FTP Toolkit* DB Loader* Monitoring GUI Utility Functions and Operators (Bloom Filter, Scheduler, …) Acquire Analyze Act TEDA Application Framework FileWritter
  14. 14. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 14 Categorize Example architecture – real-time campaigns B D A F G Real-time scoring, classification, detection and action High performance historical analysis High performance unstructured data analysis Discovery analysis Model Based Predictive Analytics Visualize, explore, investigate, search and report Take action on analytics E C Identify Score, Decide Event Execution Outcome Optimization Model Creation Policy Mgmt. Simulation Ad-hoc Queries Reports Dashboards Mapping Search, Pattern Matching, Quantitative, Qualitative PDA Big Insights Analytics Engine Hadoop Standardize Capture Changes Deduplicate Identity Resolution Batch Data TNF SourceWorks Juniper Networks Prediction / Policy Engine Open API Streaming Engine Historical Data Models Deploy Model Deploy Model IBM Campaign Management Worklight- based Mobile APP Actions . . . DPI PCMD External Data DataRepositoriesContinuousFeed Sources Streaming Data Reports & Dashboards G D CB E A F AAP Capabilities Customer Data
  15. 15. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 15 Continuous and speedy analysis in context for insurance Real-time telematic analysis: Create real-time dashboards of behaviors such as car speed and location to automatically adjust risk scores. Speedy fraud detection: Receive incident reports as they happen and immediately feed into claims processes to streamline operations. Cargo protection: Predict accidents or disasters in real time and dynamically update risk models to ensure informed underwriting. Call center optimization: Automate next best actions and increase automated responses to improve client experience, quality and performance. Insurance Increase services for clients and decrease cost and fraud Ability to model risk throughout the day vs. quarterly
  16. 16. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 16 Example architecture – big data telematics Vehicle Driver Measures Speeding Excessive acceleration Aggressive braking Cornering Driving frequency Driving distance At risk behavior Weather Temperature Time Precipitation Traffic Time Accidents Congestion Construction Changes in traffic patterns Analytics Sources Information Ingestion Information Consumption Information Governance Security and Business Continuity Management New Sources Traditional Sources Integrated Warehouse & Marts Zone (Guided Analytical Area) Integrated Warehouse & Marts Zone (Guided Analytical Area) LandingArea/AnalyticZone (Map-ReduceEngineArea) LandingArea/AnalyticZone (Map-ReduceEngineArea) Exploration Zone (Discovery / Sandbox Area) Exploration Zone (Discovery / Sandbox Area)Take action (theft, claim, accident) New Insights & Customer Opportunity Enrichment, staging, and archive Result storage and analytical access Shared Operational Information Zone Master Data Hubs Reference Data Hubs Activity Hubs Content Repository InformationVirtualization Shared Historical Information Zone Transformation Engines Streams Engine Visualization, Data Mining & Exploration Visualization, Data Mining & Exploration User Reports & Dashboards User Reports & Dashboards Accelerators & Application Frameworks Accelerators & Application Frameworks User Guided Applications & Advanced Analytics User Guided Applications & Advanced Analytics Collaboration & Insight Engines Collaboration & Insight Engines Pricing Fraud Customer Insight New Products Underwriting NBA Customer Renewal & Acquisition Cross-Sell Up-Sell
  17. 17. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 17 Continuous and speedy analysis in context for energy and utilities Outage detection and prediction: Monitor grid/plant elements and networks and rapidly predict and analyze data to detect grid/plant outages. Load shedding: Monitor and run powerful real-time analytics on data from smart meters and sensors. Condition based maintenance: Identify assets that are likely to fail in the near term or require maintenance or operational changes. Take action preemptively to control or repair equipment. Smarter Analytics: Run extremely powerful analytics from smart meters, satellite imagery feeds and weather forecasts for price fluctuation forecasting, energy trading insights and more. Energy and Utilities Optimize energy usage and reduce outages Pacific Northwest smart grid increases grid efficiency and reliability through system self-monitoring and feedback
  18. 18. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 18 Work Orders Asset Status Inventory Planning Asset Management Enterprise Asset Management Asset Master Work History Example architecture – smart grid Data Consolidation Operations Guidance Capital Planning Operational Systems Real-time AnalyticsUnstructured Structured ETL/ELTandTxNReplication Historian Structured Data Pure Data for Analytics Big Insights Information Server Data in motion Data at rest Data in many forms RCM SCADA Condition Monitoring Trading Load Forecast Demand Response Operations Logs GIS Test and Inspection Bulletins PMU/PDC Weather/ Environment EAM Data Warehouse InfoSphere Streams iLog CPLEX ACQUIRE ANALYZE ACT Optimization Predictive Analytics Informix TimeSeries High Volume / High Velocity Events Scoring Models Aggregated Streaming Data Raw and composite measurements and events Control Signals Mathematical Optimization Constraints and Rule Definition Presentation: KPIs, Dashboards, and Drill-downs Business Analytics Statistical Analytics Decision Mgt Orchestration and Integration Pre and Post Processing Analytic Data Store Predictive Maintenance and Quality (PMQ) Geo- Spatial KPIs and Integrated Dash- boards Search/ Discovery Information Consolidation and Situational Awareness Intelligent Operations Center (IOC/IOW) Resource AllocationCorrelation and Optimization
  19. 19. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 19 Shift from queries to real-time insight in context Ask Query Ask a question Find the data Analyze Store the data Is the analysis helpful? ??? Traditional Analytics Real-Time Analytics Fast Context Aware Analytics
  20. 20. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less Get the PDF: https://www14.software.ibm.com/webapp/iwm/web/signup.do?source=sw-infomg Chapter 1: Big Data at Rest and in Motion Chapter 2: In-Motion Use Cases Chapter 3: Program, Framework, or Platform Chapter 4: InfoSphere Streams Chapter 5: The InfoSphere Streams Ecosystem Chapter 6: Getting Started Appendix: Resources and References
  21. 21. © 2014 IBM Corporation Analyze More, Speed Actions, Store Less 21 Explore more on Stream Computing  InfoSphere Streams product website  IBM Context-Aware Stream Computing webpage  IBM Context-Aware Stream Computing on Big Data Hub  InfoSphere Streams developerWorks community  InfoSphere Streams Developer Community  InfoSphere Streams data sheet  InfoSphere Streams for industry alignment webpage Kimberly Madia @madiakc Avadhoot (Avi) Patwardhan @avi_patwardhan

×