Download to read offline
Uploaded byPaul Boal
Why You Should Be Using IoT Technologies for More Than Just IoT
The document discusses the Internet of Things (IoT) and its applications beyond traditional use, emphasizing the role of various technologies in enhancing business solutions. It outlines different architectural approaches for integrating IoT in business applications, data warehousing, and analytics, while debunking common myths associated with new technologies. The author, Paul Boal, advocates for real-time data processing and distributed systems to improve operational efficiency and decision-making.
More Related Content
Similar to Why You Should Be Using IoT Technologies for More Than Just IoT
Why You Should Be Using IoT Technologies for More Than Just IoT
- 1. Why You ShouldBe Using IoT Technologies for More Than Just IoT October 2017 – Paul Boal, VP of Delivery – @paulboal
- 2. What is theInternet of Things? The Internet of things (IoT) is the inter-networking of physical devices, vehicles, buildings, and other items—embedded with electronics, software, sensors, actuators, and network connectivity that enable these objects to collect and exchange data. https://en.wikipedia.org/wiki/Internet_of_things 2
- 3.
- 4.
- 5.
- 6. Wearable Devices inCider (TheApple II Magazine) October 1986 6
- 7.
- 8.
- 9. Technologies Making IoTPossible? • Physical Improvements in Electronics • Miniaturization of sensors • Low power networking (BLE, Zigbee, NFC) • Information Processing Improvements • Big data and NoSQL databases • Stream processing and analytics • Distributed processing and cloud computing • Microservices architecture 9
- 10.
- 11. Presentation / GUI Tier ApplicationLogic Tier Data Tier Classic Architectures Worth Reviewing • Business Applications • Batch EDI • Data Warehousing 11
- 12. What Architecture ShouldI Use? Round 1 12
- 13. Presentation / GUI Tier ApplicationLogic Tier Data Tier Business Applications • Round 1: You have a business application that allows business users to manage customer transactions as they go through their engagement and purchasing experience. Examples: • Web storefront • Point of sale system • Electronic health system • Utility billing system 13 MySQL, SQL Server, Oracle Java, .NET, Python HTML, Swift, JavaScript
- 14. Business Applications –N-Tier or Microservices Architectural Advantages • Clear segregation of duties • Centralized storage of data • Reusability of application logic • Create customized interfaces Enhanced with IoT • Pushing logic to the edges allows them to respond to unexpected conditions. • Streaming data allows the database to become an event communication layer as well as a storage layer. • Using a non-relational database increases the flexibility in future enhancements. 14
- 15.
- 16. Examples of IoT-Techin Business Applications Document-store database allowing flexible schema evolution. Streaming allows all applications to be notified when changes occur. Think of business users as edge nodes in the system Users and the system behave asynchronously, notifying each other when they make decisions or have information to share rather than following a fixed workflow. 16
- 17. 17 ”You’re a participant,not a user” Donald Farmer @DonaldTreeHive
- 18. What Architecture ShouldI Use? Round 2 18
- 19. Information Exchange • Round2: You have a business partner with whom you need to be exchanging information about products, services, customer verification, inventory levels, service availability, and sales transactions: • Health Insurance Member Eligibility • Billing Transactions • Healthcare Orders and Prescriptions • Funds Transfer • Order Fulfillment 19
- 20. Batch EDI /Integration Architectural Advantages • Message encapsulation and standardization • Simple text-based data exchange • Auditability and confirmation of transactions Enhanced with IoT • Generation of EDI messages during business processes allows for real-time quality assurance and feedback to operations. • Document store databases alleviate the impedance mismatch between RDBS and messaging. 20
- 21. Examples of IoT-Techin Integration 21 Systems of record publish all updates to streams. Integrated outputs can be produced at multiple intervals. Extracts leverage a collection of shared and some independent transformations. Flume
- 22. What Architecture ShouldI Use? Round 3 22
- 23. Data Warehousing andBusiness Intelligence • Round 3: You have an analysis and reporting system that takes information from several source systems and external data, merges and summarizes that information, identifies key metrics, and makes that information available to users and other downstream processes. Examples: • Operational Data Store • Data Warehouse • Data Extracts / Data Integration • Business Intelligence 23 Teradata, Oracle, SQL Server Informatica, IBM, IBI Business Objects, Microstrategy, Tableau
- 24. Data Warehousing andAnalytics Architectural Advantages • Data quality controls • Metadata management • Data standardization • Value through data integration • Business view of information • Self-service data access and reporting tools Enhanced with IoT • Switch to streaming data integration to minimize outages and hours of batch processing. • Use streaming data quality checks to send near real-time feedback to business users and improve data quality same-day. • Use messaging to feed detail tables and aggregations simultaneously rather than serially. • Use graph database to understand complex business models like networked relationships. 24
- 25. Examples of IoT-Techin DW/BI Business application streams data to Kafka Data warehouse modeled as a knowledge graph to capture complex relationships between transactions Streaming data quality checks give real-time feedback to improve business processes Graph analysis leads to easier root cause analysis 25
- 26. 26 ”The natural orderof the world is a graph not a spreadsheet.” Kirk Borne @KirkDBorne
- 27. Many Sources ofthe Truth? 27 The best way to build a data warehouse was to create a single database to control a single version of the truth. Ubiquitous distributed processing, flexible data stores, and standard communication protocols could allow a collection of analytics to be reliably shared without having to put them all in a monolithic, specially built database.
- 28. Fight the Myththat New = Hard Saying “this solution doesn’t need that new technology” promotes the myth that “new technology” is necessarily harder and more expensive. 28
- 29. Top Myths • Thetransaction overhead for real- time / streaming is too high. • NoSQL and Big Data is only for unstructured data. • Businesses want well-defined workflows they can control • There isn’t enough expertise around to build this way. • Distributed processing and databases make this irrelevant. • NoSQL is straightforward to work with in un- and structured forms. • Managers want well-defined workflows. User do not. • Open Source and cloud trends make this easy to learn and growing rapidly. 29
- 30. Thank You! Paul Boal @paulboal •Healthcare data and analytics solutions • Big data, IoT, and advanced analytics • Data strategy and data governance • Drive change through data insights 30 VP Delivery http://amitechsolutions.com @AmitechSolution
- 31.
Editor's Notes
- #16 Using a non-relational database, if we need to add a new attribute on the fly, we can create a full-fledged addition just by accepting and storing the new data. It gets fully integrated into the data model and linked to related attributes automatically. It doesn’t break anything and is immediately useful as soon as I start using it, without the need for a backfill from the producer of the new data element. Someone added that field because they need it. They felt they needed it because the capability wasn’t readily apparent in the application. If it was there and not apparent, I’d argue that is a governance and data management issue. “The right way should be the most visible and easiest way”
- #17 Turn the Ship Around – David Marquet Rather than having a centralized a necessarily infallible and purpose built captain (process) telling all the crew (business actors) what to do, each of the actors declares their intent to the system: ”I intend to order 100 boxes” I need to be prepared to receive 100 boxes I need to be prepared to pay for 100 boxes Until someone shouts “cancel that!” everyone behaves proactively as if the event will indeed happen
- #21 One of the things this enables is the ability to do “mock” transactions. For example, one of the big fights in healthcare between providers and payers is the automatic rejection and payment rates. Supply chain – real-time feedback on the Requisition / PO / Invoice / Payment process