IBM Watson IoT Portfolio
A quick look at our Toolbox for IoT
Branko Tadić, Enterprise Solution Consultant, IBM Cloud CEE branko.tadic@rs.ibm.com IBM’s contribution to the World
eBusiness
2 3
IBM contributions to Open Source: 18 years & counting
1999 - 2001 2002 - 2005 2006 - 2009 2010 - current https://developer.ibm.com/code/openprojects/
▪Contributions for Linux on Power, ▪IBM forms Linux ▪Linux contributions to ▪Linux contributions to kvm, oVirt, & Technology Center scalability (8-way+), usability, security certifications support Open Virtualization Alliance reliability (stress testing, ▪Contributes to Apache Shindig ▪Leads Apache projects defect mgmt, doc) ▪Leads Apache projects Tuscany (SCA Xerces, Xalan, SOAP standard), OpenJPA, UIMA ▪Supports Apache Hadoop (Big Data) ▪Leads Apache projects in – part of IBM BigInsights ▪Starts ICU project Web Services ▪Contributes to Eclipse Higgins ▪Eclipse: Orion (web-based tooling), ▪Creates OSI-approved ▪Leads Eclipse projects ▪Partners with Zend PHP Lyo (OSLC), Paho (M2M protocols) IBM Public License GEF (editing), EMF (modeling), XSD/UML2 ▪Announces OpenJDK involvement ▪Strategic participation ▪Accessibility code to Firefox (XML Schema), Hyades ▪Contributes to Apache Cordova (fka in Mozilla (testing), Visual Editor, ▪IBM starts OpenAjax Alliance and PhoneGap) (mobile app framework) AspectJ, Equinox (OSGi ▪IBM becomes founding joins Dojo Foundation bundles) ▪Starts Dojo Maqetta (RIA tooling) member of OSDL ▪Eclipse Foundation, Inc. ▪IBM joins OpenOffice.org & creates ▪Leads Apache OpenOffice ▪Founder of Eclipse.org becomes independent ODF Toolkit Union & Eclipse Consortium ▪OpenStack: platinum sponsor of ▪Pledged 500 patents to ▪IBM joins Open Health Tools, merging independent Foundation; over 140 ▪Creates internal open source code from Eclipse OHF contributors bazaar using OSS methodology ▪Contributes to Mozilla Bespin (web ▪Increase OSS projects & visibility at ▪Starts Apache Derby editor) & WebKit (browser engine) database, supports JazzHub and GitHub Geronimo app server ▪Lead Apache Aries (OSGi Enterprise) ▪Contributes to Cloud Foundry More than 1000 IBM developers IBM leads 80+ OSS projects IBM contributes to 150+ involved in OSS projects OSS projects IBM Watson IoT Platform Make sense of data to optimize operations, manage assets, rethink products and services, and transform customer experience.
Connect Connect and manage devices, networks and gateways. Analytics Gain insights from information using real-time streaming as well as machine learning and cognitive analytics in the ANALYTICS cloud and at the edge. Risk Management Visualize the IoT landscape, manage risk, and build trusted INFORMATION sources of IoT data with innovative technology such as CONNECTION blockchain. MANAGEMENT Information Management Integrate information, structured and unstructured, from devices, people, the weather and the world around us.
4 IBM Watson IoT Platform Connect
Connect your devices, equipment, and workforce to ANALYTICS RISK MANAGEMENT gain a new level of insight into •Real-time •Proactive Protection •Machine Learning •Security Analytics your business •Cognitive •Anomaly detection •Edge
• Secure Connectivity INFORMATION CONNECT MANAGEMENT • Device Management •Store & Archive •Transform & Integrate • Visualization •Augment •Weather
5 5 IBM Watson IoT Platform - Connect and manage your IoT devices & gateways
• Open standards based communications (MQTT, HTTPS) • Secure communication (TLS) • Globally scalable starting with a single device • Fully integrated Gateway support • Broad and growing device ecosystem
6 IBM Watson IoT Platform Integrated device management
• Manage via dashboard or Location Update programmatic APIs Factory • Action device management Reset functions on thousands of devices Firmware at a time Reboot Download • Create your own custom device management commands Firmware Update
7 IBM Watson IoT Platform Information Management
Identify, aggregate, and transform data from your IoT ANALYTICS RISK MANAGEMENT sources into asset-based data •Real-time •Proactive Protection •Machine Learning •Security Analytics structures. •Cognitive •Anomaly detection • Store and Archive •Edge • Transform and Integrate CONNECT INFORMATION •Secure connectivity •Device management MANAGEMENT • Augment with Weather & •Visualization Unstructured data
8 8 Information Management
• Built in last event cache Always have access to the last reading whether device is on or offline
• Fully managed NoSQL JSON document store built for high integrity and high performance
• Internet scale buffering between the IoT Platform and your chosen storage service, IBM Message Hub® with bridge to other Bluemix services, such as IBM Object Store
9 Information Management New Services Capabilities
Device Abstraction
Define your own APIs to insulate applications from variability across device types, sensor models, variants and versions
Example: Different models and brands of temperature sensor represented by a single common API
Aggregation into Things
Aggregate multiple devices into logical objects so they can be managed as a single Thing
Example: Several different sensors represented as a single boiler object
10 IBM Watson IoT Platform Risk Management
Manage risk and gather insights across your entire IoT landscape. ANALYTICS •Real-time RISK •Machine Learning MANAGEMENT • Proactive Protection •Cognitive •Edge • Security Analytics • CONNECT INFORMATION Anomaly Detection •Secure connectivity MANAGEMENT •Device management •Store & Archive •Visualization •Transform & Integrate •Augment •Weather
11 11 Risk Management & Policy Dashboard Your single perspective on IoT risk exposure
• Implement and accumulate reusable checks to identify device compromise and malicious events • Protect against threats to the IoT environment with blacklists, whitelists and device behaviour thresholds • Maintain platform resilience by acting on alerts automatically
12 IBM Watson IoT Platform Analytics
Leverage a host of cutting edge cognitive tools to gain a deeper understanding of your structured and unstructured data. RISK MANAGEMENT ANALYTICS •Proactive Protection •Security Analytics • Real-time •Anomaly detection • Machine Learning • Cognitive Natural Language, CONNECT INFORMATION •Secure connectivity MANAGEMENT Text, Video and Image •Device management •Store & Archive •Visualization •Transform & Integrate Analytics, Machine Learning •Augment •Weather • Edge
13 13 IBM Watson IoT Platform - Analytics Add cognitive capabilities to IoT data to produce new insights and intelligence Real-time Analytics Cognitive • Rule-based analytics and actions built in • Watson API families allow easy integration to the platform of cognitive analytics into IoT apps • Easy to use interfaces that drive • Natural human interaction, learning from automation of prescribed actions historical data, analysis of image and contextual data sources, analytics, and insights Machine Learning • Integrated IBM Predictive Maintenance Edge Analytics and Quality and Watson Machine Learning • Single click deploy of RTI rules from services Cloud to Edge • Visibility of usage and operating • Open SDK extending gateway choice conditions based on environment • Analysis of device data using IBM Data Science Experience to build custom analytics for your assets 14 Watson IoT Platform Analytics: Real-Time Insights
• Rules and action oriented analytics, built in to the platform • Business user oriented interface • Drive automation to take appropriate, prescribed actions Watson IoT Platform: Edge Analytics Reduce data feeds, make local decisions, work disconnected
• Single click deploy of RTI rules from cloud to Edge
• New open SDK extending gateway choice How is IBM Watson IoT Platform different?
Industry Leading Analytics Unmatched Scale and Scope Most Trusted IoT Platform
Enterprise-ready • Watson-inside – machine • Global data centers – 40+ data • Device neutral – IBM does not components to learning and cognitive centers across the globe compete with its sensor, gateway, connect, secure, network, and processor partners provide data insight, • Industry models – deep, • Low latency and high assemble and industry-specific analytics models throughput at enterprise scale • Built on open standards manage IoT Applications • Third party data sources – • Hybrid delivery form factors... • Data neutral – IBM’s business model leading the industry at partnering public cloud, dedicated cloud, on does not depend on owning its with outside data providers (e.g. premise customer’s data Weather Company) • Bluemix and Softlayer – built to • Privacy protection and access control • Industry Integrations – easily work on IBM’s core cloud push and pull data from leading offerings but also deliver the • Platform of Platforms – IBM is industry solutions, both IBM’s and transactional scale required by the committed to integrating with other its partners’ new world of IoT leading platforms so customers are not forced to chose proprietary tech stacks
• IoT specific security – security micro- The Weather Company APIs and data services built specifically for IoT-based IoT Platform integrated with Blockchain solutions
17 A brief Platform Showcase Example 1 – Turn your Mobile into an IoT device
http://discover-iot.eu-gb.mybluemix.net/#/play
19 IoT Platform Starter on Bluemix IoT Platform Starter Boilerplate
21 NodeRed
22 Management Portal
23 Extensions
24 Example 2 – Quick Walkthough of IoT Starter
www.bluemix.net
25 IBM Watson IoT Industry Solutions IBM Watson IoT and Industry Innovation Enabling new business models with integrated solutions
Transform traditional business with the Insurance Transport capabilities of IoT Industrial Electronics • Drive customer relationships & Products experiences Retail • Improve operational efficiency & reduce costs Automotive Aviation • Deliver new product and business Energy & Chemical & models Utilities Petroleum
• Drive better customer engagement Healthcare • Leverage Watson for cognitive solutions
27 14 IBM Watson IoT for Automotive Enabling the next generation of connected vehicles
A Next Gen Vehicle 80% of new apps will produce more than will be distributed or 50 GB of data per hour deployed on cloud
Real-Time Deep Analytics Dynamic Map Management Road Network Dynamics
•Nanosecond level high speed •Store and analyze historical •Efficient in memory map store •High accuracy and high computing information for actionable and IDE for application scalability map matching insights development •Real-time awareness of vehicle •High performance trajectory and surrounding •Traffic sign identification and •Multiple map vendor and version data management & analytics map generation support
1528 Value-added services further differentiate our IoT for Automotive offering
Map Insights Driver Insights Vehicle Insights ▪ Personalized mobility services ▪ Store and analyze historical • Real time Contextual information for actionable information ▪ Store and analyze historical insights driving behavior and vehicle • Awareness of vehicle and usage information ▪ Optimize assets and supply surrounding chain
Capability Capability Capability Environmental awareness & Contextual Per vehicle & driver real-time awareness Store and analyze historical vehicle information condition information Highly accurate and scalable map Store and analyze historical driver and Vehicle asset information matching vehicle usage information Data integration with multiple systems of High performance road attributes such as: High speed, low latency messaging & record traffic sign, speed limit, link-node distributed cloud environment Enhance OBD capabilities with IoT Cloud network, give way Unique Agent system High performance and trajectory data 29 analysisWatson IoT Our automotive clients are among the leading industries on the edge of IoT-based transformation
PRODUCT ASSET MANAGEMENT ENGINEERING
INFORMATION CONNECT MANAGEMENT
RISK MANAGEMENT ANALYTICS
30 Watson IoT Example 3 – IoT for Automotive Experience
https://iot-for-automotive-starter-experience.mybluemix.net/
31 IBM Watson IoT for Electronics Enabling the next generation service delivery of connected products
80% of new apps <1% of data is currently used. will be distributed or More must be used for deployed on cloud optimization and prediction.
Scale Analytics Life cycle Management Client Experience
•Tens of millions of devices, on a •Store and analyze information for •Onboarding to updating – secure •Improve product & client cloud infrastructure across >44 actionable insights and pattern and efficient engagement thru connectivity and data centers awareness analysis of usage •Aftermarket service •Cost efficient and secure •Real-time for rapid awareness management, from work orders •Reduce service costs with timely / information management and resolution to work scheduling accurate information
1632 Example 4 – IoT for Electronics Experience
Demo IoT for Electronics Instance:
33 IBM Watson IoT for Insurance
Protected Lifestyle & Business Risks Assets Finances
Protected Protected Wellness / Workers Cars / Fleet Safety Contextual Health Analysis Driver Profiles Insights Telematics Mapping
Real-time Monitoring Alerts
Protected Home & Buildings Protected Homes Ships Assets & Equipment Factories Connected Environment Predictive Maintenance Warehouses Weather Authorized Access Construction Sites Construction Location Commercial Office Crime Carriers are transforming their historic risk assessment 34 models by proactively mitigating risk through real time alerts and analytics IoT for Insurance tailored for Proactive Protection Aggregation – correlates data, Transformation applies against intelligent rules Collect and Normalize Data Industry specific Analytics for potential Hazards
Simple Home Devices Insurance Industry • Water • Water Intrusion • Home/Auto/Workers Auto Data • Gas exposure • Overexertion compensation • Driver Behavior • • Man Down Policy holder alerts • GPS • • Ice Slip/fall Device utilizations • Wearables • Heat or Cold stress Claims analysis • Body Temp • Fraud analysis • Blood Sugar Predictive • Risk dashboards • • Vitals Avoidance of issues • Potential Black Mold Environment Sensing • Noise exposure over time • Weather • High risk areas Other Industry • Light • Fall Prevention • Safer Workplace • Noise • Employee alerts • Gas Cognitive • Fall prevention • Supervisor dashboards Equipment • Alertness • Injury prevention • Diagnostics • Dehydration • Elderly Care • Fatigue • Use of safety equip • Device and External data Alerts • Standardized formatting • Connect data sets • Text • Annotation/augmentation • Real Time • Email • Data published • Hazards / Insurance Risk • Emergency
Devices / Raw Data Shields (1..N) Industry Analytics 35 DataIBM Confidential Providers 35 © 2016 IBM Corporation IBM Watson IoT for Insurance Safer Workplace Body-worn and environmental sensor stack: The “Guardian Angel” App - Movement - Heart rate • Mobile as-a Gateway - Body Temperature • Linkage to guards libraries - EEG - Location • Shields subscription mgmt - Air quality • Sensor-stack admin - Noise • Messaging/Notifications
Shield Library Stream analytics components reflecting a concrete hazard(e.g., Node-Red flows) Fall | alertness | dehydration | fatigue …
Safelet Runtime Data Store Node JS, BlueMix rules Cloudant IoT for Insurance Safer Workplace 36 36 Example 5 – IoT for Insurance
Employee Wellness and Safety Demo https://www.youtube.com/watch?v=8-j26pA9Wrg
37 IBM Watson IoT 4 Telco
• How do I reassert relevance in B2B Markets? • How do I find new things to sell? • Services How do I partner to deliver value? • Innovation How do I participate in the API economy? • How do I make my offerings ‘cognitive’?
• How do I manage my fleet, network, masts? • How do I manage my employees? • How do I manage my stores? My sites? Enterprise • How do I secure my enterprise?
• How do I manage signalling? • How do I compete against low power competitors? • How do I secure IoT? Network • How do I differentiate?
• The Watson IoT Platform helps to drive “Services Innovation” in particular, and the focus of most discussions with telecom service providers is in this domain. The others are important considerations, however, for our clients.
38 Introducing Mobile Asset Optimization from Vodafone and IBM
Data Capture Capture the location data from all your assets
Connect Transfer the data to Vodafone’s and IBM’s cloud
Analytics Leverage intelligent analytics to provide insights
Tracking Track your assets to ensure seamless operations
Notifications Receive notifications through SMS for extreme cases
Decisions Make informed decisions to minimize impact on overall business
Result Make data-driven logistics decisions How Mobile Asset Optimization works
Device will Data integrates with report to a M2M platform manages device data, and integrates with Customer’s asset is fitted advanced analytics to central server IBM Watson IoT Platform via. a Cloud to Cloud with a tracking device provide predictive via Vodafone Connector insights Network
Customer Alarm management Customer Asset Vodafone M2M Platform Reporting and maps
KPI’s and notifications Business logic API
Solution offered as a Data storage Weather impacts managed IBM Cloud to The customer Tracking cloud Cloud Connector supplies this part - Device Availability Prediction service Vodafone does Risk/Loss Zone the rest Watson IoT Device integration detection Platform and management Lifecycle optimization
MAO Dashboard Containers and trailers Electric bicycles
• Customer: Large UK based SME • Customer: Netherlands manufacturer of electric bicycles. • Problem: Locating container fleet . Regulatory fines • Problem: Low security but high-value electric bike, low when trailer maintenance goes over-schedule market differentiation • Solution: CalAmp ATU620 fitted to trailers and • Solution: CalAmp TTU1220 engineered into the bicycle containers, reporting once a day, Device lifecycle: cowling and wired to the bicycle battery, movement 18months before battery change. ROI: 1.2 years sensed reporting . ROI: 2 years • Analytics: battery life-cycle, route cycle times, weather • Analytics: usage patterns, maintenance schedules, loss impact zones
Specialised stillages for Coffee machines closed-loop circulation
• Customer: Multi-national manufacturing vehicle • Customer: Service/maintenance provider in New windscreens Zealand • Problem: Poor flow of transit cages causing backlogs at • Problem: Machines are moved and cannot be factory. 4000 cages per year lost. maintained by visiting technicians or mistakenly disposed • Solution: CalAmp ATU620 fitted to each ‘stillage’, of at end of lease. movement sensing, Device lifecycle: 10 months before • Solution: Zelitron ZLT-AT-11 inserted inside the machine, battery change. ROI: 3 years reporting once a day on Cell-ID and monitoring for • Analytics: battery life-cycle, loss zones, availability. tamper. ROI: 1 year • Analytics: maintenance schedules, loss zones IBM Watson IoT for Retail
BUILD SMARTER DELIVER A SMARTER MERCHANDISING AND SHOPPING EXPERIENCE SUPPLY NETWORKS
• Smart Shelf • Smart POS • RFID Solutions • Monitoring & Tracking • Smart Trolleys • People counter • Payment systems • Scanning and weight control • Queue management • Anti-theft systems • Security & surveillance • Vending and reverse vending machines
DRIVE SMARTER OPERATIONS
42 Open ecosystem & partnership strategy extend IBM Watson IoT platform
Derive IoT value on the Cloud through strong industry partnerships and open ecosystem Wide variety of Examples: supported devices ✓ Self Service ✓ Open ecosystem ✓ Simple tutorials ✓ Connect in moments
https://developer.ibm.com/recipes/tutorials/ 43 Watson IoT Platform meets Machine Learning
Hands on Lab
Engage Machine Learning for detecting anomalous behaviors of Things
Branko Tadić, Enterprise Solution Consultant, IBM Cloud CEE [email protected] Key Links and Prerequisites Detailed instructions (Recipe) for the Lab: ibm.co/2bwi5zj
Bluemix PaaS home: bluemix.NET (register for a free 30 day trial accnt)
IBM Data Science Experience (DSX): datascience.ibm.com (register for a free 30 day trial accnt)
Git and Maven installed
JDK installed ☺ 45 Basic Terms •What is Machine Learning? •Machine learning is a method of data analysis that automates analytical model building. Using algorithms that iteratively learn from data, machine learning allows computers to find hidden insights without being explicitly programmed where to look
•What is Predictive Analytics? •Predictive analytics encompasses a variety of statistical techniques from predictive modeling, machine learning, and data mining that analyze current and historical facts to make predictions about future
•Types of ML Algorithms • Supervised learning • Unsupervised learning • Reinforcement learning
46 Engage Streams to detect Anomalies
• Anomaly Detector operator reports anomalies with the pattern of the incoming IoT data • The operator maintains a recent history of the input time series, which is referred to as the reference pattern • The operator compares the current pattern with the reference pattern and generates a score 47 Scenario This recipe explains how one can integrate IBM Watson Machine Learning service with IBM Watson IoT Platform to predict a temperature change before it hits the danger zone.
48 Ingredients • Temperature sensor simulator (Java source available on GitHub, https://github.com/ibm-messaging/iot-predictive-analytics-samples.git) • Watson IoT Platform instance, on Bluemix • Apache Spark instance, on Bluemix • Watson Machine Learning instance, on Bluemix • Object Storage Service, on Bluemix • Machine Learning Streaming Predictive Analytics Model, available on GitHub, https://github.com/ibm-watson-iot/predictive-analytics- samples/raw/master/SPSSModel/nocycle20rebuid50.str
49
Architecture Simulator
50 So, let’s start!
51 Thank You
52