building connected enterprises - amazon s3...azure iot reference architecture iot. analytics....
TRANSCRIPT
IoT. Analytics. Mobile. Cloud
December 2015
Building Connected EnterprisesUsing Microsoft Azure IoT Suite
IoT. Analytics. Mobile. Cloud
IoT. Analytics. Mobile. CloudWhy ‘Internet of Things’
• The term “Internet of Things” isn’t new.
• Almost 20 years ago, MIT professors described a world where “things” (devices or sensors) are connected and able to share data.
• Connecting devices to IT systems & Sharing data is only the first step.
• Data coming from these devices and sensors, is valuable only when it provides business insights that were previously out of reach. These “invaluable insights” enabled by ‘harnessing and analyzing the data’ from these ‘connected devices’ are what the Internet of Things is all about.
• Connected Enterprises is the end goal of a successful IoT implementation.
Connectivity Data AnalyticsThings
IoT. Analytics. Mobile. Cloud
IoT. Analytics. Mobile. Cloud
IoT. Analytics. Mobile. Cloud
Why is IoT relevant today, than ever?
Hardware costs are falling.
Connectivity is proliferating.
Cloud solutions offer lower costs, scale, and flexibility.
Software is more advanced than ever.
IoT. Analytics. Mobile. CloudEg: Smart Meters & Energy Management solution
Business Objectives
• Real-time visibility & dashboard of water consumption
• Remote Asset management & maintenance
• Cost-efficient Smart-Metering & Meter Data management
• Enterprise-wide Mobile application for Energy Utilities
IoT. Analytics. Mobile. CloudSmart-meter & Utilities solutionDenver, US
IoT Solution• Give real-time visibility of Water consumption, Meter health, Water leakage/theft, hourly consumption data, etc.
• A highly cost-efficient & rapid deployment 'Internet of Things' solution for this Water Meter Manufacturing company.
• Operations Volume: 500,000 water meters; 200 Million + records; less-than a second response times
• Total cost of Solution development: USD 175,000 – 200,000– Proof of Concept: 6 weeks; USD 20,000
– Pilot Solution development: 10-12 weeks; USD 45,000 – 60,000
– Full-scale deployment: 6-8 Months; USD 110,000 – 130,000
– Ongoing development, support, enhancements, upgrades
IoT. Analytics. Mobile. CloudWater Utilities management solution
IoT. Analytics. Mobile. Cloud
IoT. Analytics. Mobile. Cloud
IoT. Analytics. Mobile. CloudWhy Microsoft Azure for IoT implementations
1. Azure starts by building on the infrastructure you already have in place, with your familiar devices and services, then incorporating the required technology to help you use data to create insights and make more informed business decisions.Azure IoT Suite starts from where you are today—whether you’re starting or ready to scale your years of investments in existing IoT scenarios.
2. Data from ‘your Things’ will have different formats, values, retention requirements, and traffic patterns. It will come from different sources (devices, services, etc) and across different protocols. Microsoft Azure IoT automates the reception of that data, by providing a framework for the data to ingress and be processed - through filters, rules, triggers, etc.
3. In addition, data in the typical Internet of Things scenarios is large—too large & Complex, that it becomes very costly to Consume and extremely difficult to process using typical on-premises database management and processing applications.Microsoft Azure IoT helps possessing data with a set of tools, engines, and scalable architecture model, that evolves as your business evolves. Azure makes your data contextual, combining it with many other assets, sources, and datasets.
IoT. Analytics. Mobile. CloudAccelerate time to value with preconfigured solutions
Modify existing rules and alerts
Fine-tuned to specific assets and processes
Integrate with back-end systems
Highly visual for your real-time operational data
Get started in minutes
Add your devices and begin tailor to your needs
IoT. Analytics. Mobile. CloudIndustry Examples
• When you use cloud-based solutions for storage and analysis, you can combine data from multiple sources without worrying about capacity constraints or the significant costs that might result from building out your on-premises infrastructure.
a. For example, the equipment involved in mining, moving, refining, and selling petroleum is expensive and rugged, and comes from hundreds of manufacturers. Enhanced by the Internet of Things (IoT), Rockwell Automation is extending its systems that monitor these valuable capital assets and use that data for predictive and even preventive maintenance. The solutions have the potential to transform the petroleum supply chain and produce bottom-line results in global productivity that could ultimately pay off at the pump.
b. ThyssenKrupp Elevator wanted to gain a competitive edge by focusing on what matters most to its customers in buildings the world over: reliability. Drawing on the potential of the Internet of Things (IoT) by connecting its elevators to the cloud, gathering data from its sensors and systems, and transforming that data into valuable business intelligence, ThyssenKrupp is vastly improving operations, and offering something its competitors do not: predictive and even preemptive maintenance.
IoT. Analytics. Mobile. CloudAzure IoT services
IoT. Analytics. Mobile. CloudMicrosoft Azure IoT servicesProducers Connect Devices Mine Data Take Action
Event Hubs (Service Bus)
SQL Database Machine Learning Azure Websites
Heterogeneousclient agents
Table/Blob Storage HD Insight Mobile Services
External Data Sources
DocumentDB Stream Analytics Notification Hubs
External Data Sources
Cloud Services Power BI
Microsoft delivers technologies to enable large-scale IoT solutions— with Cloud, Networks/Gateways, Heterogeneous device support, Systems capabilities, and Data Analytics.
External Services
{ }
IoT. Analytics. Mobile. CloudEg: Real-time Delivery Tracking
IoT Solution• Give real-time visibility of complete Supply chain, their Assets, Drivers and Trucks, by building a cloud-based
Enterprise Mobility platform.
• One Integrated application, to create a business-transformative 'Internet of Things' solution for this Logistics company.
• It integrates their drivers' mobile devices with their ERP, Azure cloud, bluetooth printers & temperature sensors inside refrigerators on the trucks, for real-time operations tracking.
• Operations Volume: 50,000 orders/day; 300+ Mobile workforce; Millions of daily Transactions
• Total cost of Solution development (approx): USD 85,000 – 105,000– Proof of Concept: 2 weeks; USD 5,000 - 7,000
– Pilot Solution development: 4-6 weeks; USD 20,000 – 30,000
– Full-scale deployment: 12-15 weeks; USD 60,000 – 70,000
– Ongoing development, support, enhancements, upgrades
IoT. Analytics. Mobile. CloudEg: IoT Solution Architecture
IoT. Analytics. Mobile. CloudA. Connect IoT Devices
You can connect all your devices to the cloud, receive data
at scale from those devices and manage the authorisation
and throttling of those devices.
Easy provisioning of capacity to process events from
millions of devices while preserving event order on a per
device basis, to run your applications on many device
platforms.
IoT. Analytics. Mobile. CloudB. Real time monitoring
You can analyse millions of events per second in the cloud
and rapidly develop and deploy a real-time monitoring
solution that enables actionable insights from the data sent
by the devices and sensors in your infrastructure.
IoT. Analytics. Mobile. CloudC. Anomaly detection in real time
When you are monitoring the event stream from your
connected devices, you can forward events to a machine
learning algorithm which can identify anomalies in the
patterns of data that might indicate a problem in your
business processes or infrastructure.
You can then configure a trigger for an alert on a real time
dashboard with a notification to your administrators.
Administrators can use these alerts and notifications to fix
problems or to preemptively prevent a problem from
occurring.
IoT. Analytics. Mobile. CloudAzure IoT Reference Architecture
Solution PortalProvisioning API
Device Registry Store
Stream Event Processor
Analytics/ Machine Learning
Data Visualization & Presentation
Device State Store
GatewayStorage
IP capable devices
Existing IoT devices
Low power devices
Presentation Device and Event ProcessingData
TransportDevices and
Data Sources
Event Hubs
Azure IoT Hub
Agent
Agent
Agent
IoT. Analytics. Mobile. CloudIoT & Predictive Analytics – examples
IoT. Analytics. Mobile. CloudUse Case 1 – Predicting Consumption
What we did
Finding patterns and behaviors in data.
Replacing human written code with supplying data
Divide historical Data into 2 parts, training and testing
Training data is used with different statistical models for understanding patterns
Testing data is used to execution or scoring model to test and get the result
The experiment can be published as Web service to integrate with applications
Customization can be done using R Script language
How AzureML helped
A fully managed service in the cloud that allows users to publish advanced analytic web services in minutes and build enterprise applications
Cloud-hosted ML solution to transform data into actionable business insights.
Using this approach, utility companies “can now connect smart meters and software to forecast consumption, reduce leaks and monitor temperatures and Battery voltage to improve efficiencies
IoT. Analytics. Mobile. CloudUse Case 2 – Customer Retention
What?
• For every business, some customer churn – leave
• What if we can tell ahead of time which customers are going to leave?
How?
o Machine Learning Algorithm can “learn” from previous history transaction patterns
o Use the trained model to make predictions
o Target customers most likely to leave
Customer Success Story
o Top 20% of churn predictions are over 3x better than random pick
o Out of 200 predicted using our system ~45% retained
o Past experience without our system was ~17% retention rate
IoT. Analytics. Mobile. Cloud
Thank You
25
Thank you
Anubhav DwivediCEO, Founder – [email protected]