Realizing Business Value from Convergence of IoT + Big...
Transcript of Realizing Business Value from Convergence of IoT + Big...
Realizing Business Value from Convergence of IoT + Big Data Technologies
Aditya Thadani | Phil Andreoli
CONVERGENCE OF IOT+BIG DATA TECHNOLOGIES Agenda
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 2
Business Opportunity & Outcomes
Where does IoT (Internet of Things) fit in Business strategy for Industrial Enterprises?
What Business outcomes are Industrial Enterprises looking for from IoT?
Technical Challenges
How do Big Data technologies align with the IoT opportunity?
What are key Technical challenges in realizing these Business Outcomes?
Organizational Challenges
What are key Organizational challenges in realizing the IoT+Big Data Architecture?
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 3
• Diversified - Oil&Gas, Food processing, Dairy, Pharmaceuticals, Pulp & Paper, Mining, Textile Care, Hospitality, Food Retail, Restaurants, Healthcare, …
• Global footprint with aggressive growth in emerging markets
BUSINESS CONTEXT About Ecolab
• Where does IoT/M2M fit in business strategy for Industrial Enterprises?
• What business outcomes are we looking for?
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 4
IOT IN INDUSTRIAL ENTERPRISES Business Opportunities
2009 2010 2011 2012 2013
For the third consecutive year, Ecolab was named to Forbes’ list of “The World’s Most Innovative Companies.” … ranked 33rd out of 100 companies …
Market Differentiation through Innovation
Value Proposition based on Customer Outcomes
Innovation KPI - Vitality Index
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 5
IOT IN INDUSTRIAL ENTERPRISES Strategic Alignment - Ecolab Perspective
Opportunity to extract latent value from an Asset base – Network of sensors in Customers’ Operations environment
Key Messages
Upstream
Downstream
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 6
IOT IN INDUSTRIAL ENTERPRISES Business Opportunities - Ecolab Perspective
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 7
Differentiate in the Marketplace
Maintain price-premium against competitors through Value-added services
Demonstrate Contract value to enterprise customers through KPIs
Drive Product Innovation
Use real-world application data to refine & develop new Products & Applications
Develop sustainable solutions that reduce consumption of Scarce resources
Monetize New Information Services
Higher-value Operations Management service offerings to existing customers
Data-driven Insights to Extend service life of Capital assets
IOT IN INDUSTRIAL ENTERPRISES Business Opportunities
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 8
As IoT Economics shift:
• # of Networked Sensors is exploding
• Shifting from “Capture only What’s
needed” to “Capture in case it’s needed”
IOT CONCEPTUAL ARCHITECTURE Opportunities & Challenges
Use Sensor data for:
• Operations Monitoring (NOC-like service)
• Near-real time Predictive Analytics
• Insights for Product Innovation (RD&E)
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 9
IOT IN INDUSTRIAL ENTERPRISES Market Opportunity & Challenge
The Opportunity
The Challenge
Explosive growth in Networked devices
Our ability to make sense of the data
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 10
Big Data Platform
Volume
Millions of Sensors
Velocity
Data streaming in every minute
Variety
Diverse Sensor Apps
IOT+BIG DATA TECHNOLOGIES How Does Big Data Align with IoT Challenges?
• While in B2C, Big Data platforms were initially used to deal with Search & Social data
• In Industrial Enterprises, Big Data platforms can be leveraged to add Machine data to the mix
• What are some of the Technical Challenges in realizing these business outcomes?
• What architecture approaches can we employ to overcome these challenges?
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 11
IOT+BIG DATA ARCHITECTURE Technical Challenges
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 12
Velocity - Keeping up with High rate of IoT/M2M Data generation
Before: Captured “Only What’s Needed” from a limited # of networked devices
After: Build a Data processing platform that can ingest larger & more frequent
data transmissions from a growing # of networked devices - Capture
“Everything in case it’s Needed”
Solution Approach
Architect a platform for Schema-less processing of M2M data on Write AND Schema-
based on Read for Analytics
Switch from ETL-based processing to ELT-based processing; Relocate Transformation
workload to an MPP platform
IOT+BIG DATA ARCHITECTURE Technical Challenges
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 13
Variety - Friction between Sensor & Enterprise Data Infrastructure
Before: Sensor data lived largely on an island (or even several islands)
After: Build a Digital business that integrates the M2M data in Product
development, Marketing, Sales & Service operations
Solution Approach
Switch from “Data Pipeline or Data Supply Chain to Zone-based Data Architecture”
to bring together Sensor data & Enterprise Systems data
Leverage MPP platforms to support processing engines that span the boundary
between Schema-less M2M & Schema-based Enterprise data
IOT+BIG DATA ARCHITECTURE Technical Challenges
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 14
Data types
Enterprise App data
Machine and sensor data
Enterprise content
Social data
Image and video
Market data
Actionable Insights
Prescriptive
Predictive
Diagnostics Descriptive
Discovery Exploration
Master Data & Governance
Streaming Analytics
Deep Analytics
Data Marts
Data Warehouse
Landing Zone
Data Processing Platform MPP | SQL+NoSQL | ELT
IOT+BIG DATA ARCHITECTURE Conceptual Data Management Architecture
• What are the Organizational Challenges in realizing the IoT+Big Data Architecture?
• What approaches can we employ to overcome these barriers?
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 15
IOT+BIG DATA ARCHITECTURE Organizational Challenges
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 16
Veracity - Changes in Data Quality & Governance
Before: Consumers of data expect high quality data from Enterprise platforms
After: Different Business uses strike a different balance between latency, detail
& quality
Solution Approach
Establish distinct expectations of data quality across “zones”
Establish Data Governance policies by Zone - Landing Zone (Least Governed) Data
Marts & Analytics Zone (Locally Governed) EDW (Most Governed)
Shift focus from “Data Quality” measures to “Fit for Purpose”
Shift Data Governance emphasis from Control to Informed consumers
IOT+BIG DATA ARCHITECTURE Organizational Challenges
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 17
IOT+BIG DATA ARCHITECTURE Organizational Challenges
Value - Extracting Value from the Data
Before: We delivered Validated, Nicely-packaged, Limited Data sets to Users
After: We are asking analysts to mine “Dirty” Unstructured Data sets for
hidden gems AND turn those into product/service offerings
Solution Approach
Develop Organizational capabilities to go from
“Discover Insights Turn Insights into Products/Services Operationalize Insights”
Model after traditional Product Development processes that go from
R&D Product Development Manufacturing Engineering Product Delivery
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 18
IOT+BIG DATA ARCHITECTURE Organizational Challenges
Data Security - Protecting Customer Data
Before: Sensor data stayed on disconnected islands limiting the Attack surface
After: Companies are aggregating, storing & moving vast amounts of
operationally sensitive customer data potentially creating a far bigger
Attack surface
Solution Approach
No Silver Bullets
Combine & Adapt Security measures developed & deployed in Industrial & B2C
markets
11/25/2014 2014 MACC CONFERENCE | MAKING IT MATTER 19
Thank You
Aditya Thadani www.linkedin.com/in/adityathadani
Phil Andreoli www.linkedin.com/in/philnandreoli