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.

Data Science Popup Austin: For The Internet of Things, The More Things the Merrier

712 views

Published on

Watch talk ➟ http://bit.ly/1NJNJX9

As more and more devices are internet enabled, creating the so called “Internet of Things” (IoT), our ability to understand the nature and state of our environment becomes both more powerful, and more resilient. Much of the early focus on IoT has been on the new capabilities that connected devices can bring. However, much of the power and promise of IoT comes from the combined view that multiple, disparate, connected devices can provide. We will discuss the increased analytical potential of systems that incorporate multiple, overlapping observations of operations or behaviors. We will work through some examples that demonstrate how to benefit from the coming ubiquitous visibility. We will also describe how IoT will help build more resilient systems that degrade gracefully, rather than failing outright. This resilience will be critical as we become more dependent on the successful collaboration and coordination of disparate, connected devices.

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

Data Science Popup Austin: For The Internet of Things, The More Things the Merrier

  1. 1. DATA SCIENCE POP UP AUSTIN For the Internet of Things, the More Things the Merrier John Akred Founder & CTO, Silicon Valley Data Science BigDataAnalysis
  2. 2. DATA SCIENCE POP UP AUSTIN #datapopupaustin April 13, 2016 Galvanize, Austin Campus
  3. 3. FOR THE INTERNET OF THINGS, THE MORE THINGS THE MERRIER Data Science Pop-up in Austin John Akred, CTO 13 April 2016
  4. 4. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED. @SVDataScience5 Silicon Valley Data Science transforms business through data science and engineering.
  5. 5. @SVDataScience6 OUR TEAM
  6. 6. @SVDataScience7 more=better? https://pixabay.com/static/uploads/photo/2015/05/25/05/27/network-782707_960_720.png
  7. 7. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED.8 The Internet of Things (IoT) is the network of physical objects— devices, vehicles, buildings and other items—embedded with • electronics, • software, • sensors, and • network connectivity that enables these objects to collect and exchange data.
 WHAT MAKES A THING AN IOT THING? Source: https://en.wikipedia.org/wiki/Internet_of_Things  "Internet of Things Global Standards Initiative". ITU. Retrieved 26 June2015.
  8. 8. @SVDataScience9 A connected thing can do useful stuff for us, sometimes really useful. http://arstechnica.com/science/2016/04/er-docs-get-heart-rate-info- from-fitbit-save-patients-life/
  9. 9. @SVDataScience10 An ecosystem of connected things can do useful stuff for us too. https://jawbone.com/blog/napa- earthquake-effect-on-sleep/
  10. 10. @SVDataScience11 Value of 
 Connected Devices Single Device $100s/year Device Network 50% aggregate 
 demand
  11. 11. @SVDataScience12 Hard Failure: If the data source is broken, so is the app. Stovepipe: One-to-one relationship from data source to product. Multi-sourced: Redundancy of overlapping data sources makes your products more resilient. Graceful Degradation: If a data source breaks, there is a backup and your app continues to function. Production data services abstract the probabilistic integration of overlapping data sources. We call this model a Data Mesh. Why more = better? DATA RESILIENCY
  12. 12. @SVDataScience
  13. 13. @SVDataScience14 • Commuter rail between San Francisco and San Mateo and Santa Clara counties ~30 stations • 118 passenger cars • 60% >=30 years old • 2014 weekday ridership is 52,019 people daily • On-time performance is about 92% • No reliable real-time status information • API outage between April 5th and June 2nd
  14. 14. @SVDataScience15
  15. 15. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED. @SVDataScience16 HOW DO WE KNOW IF THE TRAIN IS ON TIME?
  16. 16. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED. @SVDataScience17 MISSING DATA? The departure data: • mostly “on time” • when significant delays occur, they often appear suddenly • often missing >10% per day • ~1 minute resolution
  17. 17. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED. @SVDataScience18 AUDIO
  18. 18. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED. @SVDataScience19 VIDEO DETECTION - FRAME ANALYSIS No Train Detected Train Detected RawImage (+ROIs) Relative Motion (+ROIs)
  19. 19. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED. @SVDataScience20 VIDEO
  20. 20. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED. @SVDataScience21 SOCIAL MEDIA AS SENSOR Baseline
  21. 21. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED. @SVDataScience22 CALAMITY DETECTION IS ACCURATE
  22. 22. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED. @SVDataScience23 20 THE MOST IMPORTANT SIGNAL Where is the train right now?
  23. 23. © 2016 SILICON VALLEY DATA SCIENCE LLC. ALL RIGHTS RESERVED. @SVDataScience24 PREDICTION MODEL: NEURAL NETWORK neural network prediction residuals previous delay residuals
  24. 24. THANK YOU Yes, we’re hiring! info@svds.com John Akred @BigDataAnalysis
  25. 25. DATA SCIENCE POP UP AUSTIN @datapopup #datapopupaustin

×