social convergence of machine learning in iiot - jeffrey jensen
TRANSCRIPT
HOUSTON │OSLO │ PALO ALTO
Social Convergence of Machine Learning in IIoT
IoT With the Best – 29 October 2016
Today’s Talk
| Copyright © 2016. Arundo Analytics. All rights reserved2
• Who am I? Who is Arundo? Why am I talking? Why this matters to you?
• Machine learning – where we are in the space?• What is “social convergence” anyway?• Some examples.
Me.
| Copyright © 2016. Arundo Analytics. All rights reserved3
• Actually an antennas/controls person (Ph.D. in electromagnetics/stochastic methods)
• Diverse software development experience• Triathlete• Previous employers:
• Intel• Toshiba• Siemens
Locations and Notable Logos
| Copyright © 2016. Arundo Analytics. All rights reserved6
Three offices to meet global growing demand of data science in oil & gas and marine industries
Ongoing work with many clients, a couple I can mention here…
A post data science, data science company
| Copyright © 2016. Arundo Analytics. All rights reserved7
In addition to building new algorithms, deploying models, and solving big data problems for customers…
.. We are beginning to ask questions about how data science is being communicated and used within companies in order to not only make an impact but also to scale the impact.
Therefore, from our learnings, we realize there are several challenges ahead with regard to implementing data science solutions…
… because of the amount of people, machines, and data that the solution must impact and the amount of coordination it will take across these aforementioned stakeholders
So – we do data science and have placed a good amount of focus on the “social convergence” (including machines) of data science solutions for big data problems.
The anticipation of the industrial internet
| Copyright © 2016. Arundo. All rights reserved12
Expects the Global IIoT to be $220Bn in 2020
Expects the Global IIoT to be $14.4Tn by 2022
Collection of data – even in Iiot
| Copyright © 2016. Arundo. All rights reserved13
Machine Learning
Images from iconfinder.com and pixabay.com
Several groups and people with different objectives
| Copyright © 2016. Arundo. All rights reserved14
Machine Learning
Images from iconfinder.com
How do we connect every source to maximize impact for all?
| Copyright © 2016. Arundo. All rights reserved15
Machine Learning
Machine Learning
Actions/decisions
Products for turning data into value
16
Data
Real-time failure predictions and
performance optimization
Real-time data
Historical asset performance
Batch data
Annotations and interaction across
assets and companies
Meta data
LiveQ
Q
DeepQ SocialQ
An analysis and collaboration of people solving around data science
| Copyright © 2016. Arundo. All rights reserved19
Our approach is unique*
20
Enabling our customers to get control of all data across equipment and sensors, on all assets - even industry wide
* Patent pending
1. System-wide and equipment agnostic deployment
2. Enhanced predictability through model-sharing*
3. Industrial network infrastructure