Welcome to Our Home. Anni Toner Managing Director The Data-Shack Home of Analytics in AME & APAC

Transcription

1 Welcome to Our Home Anni Toner Managing Director The Data-Shack Home of Analytics in AME & APAC

2 Christer Ekqvist Managing Director

3 IoT, Process & People Behaviour The Magic Combo

4 The Internet of Things

5 Why are Analytics essential to IoT? "Data is inherently dumb, it doesn't actually do anything unless you know how to use it and how to act on it, because algorithms are where the real value lies; algorithms define action, November 2015 Gartner Symposium 2015 in Barcelona, Peter Sondergaard, senior vice president and head of research at the analyst house

6 6 Making Sense of Big Data

7 IoT Information lifecycle 1. Plan Targeted outcome RIO analyze Start small, use available data 2. Acquire Sensors, controllers and actuators Gateways Edge security and manageability 3. Transport Ethernet, WiFi, Cellular, ZigBee, MQTT Minimize bandwidth consumption and solution latency Edge/Fog Solutions 7 Predict 7. Plan 6. Predi ct Analyze Act 1. Acqu ire Security 5. Act Aggregate Plan2. Transp ort Acquire 3. Aggre gate 4. Transport Anal yze 4. Aggregate Tiered data aggregation Cloud brokering Minimize data management cost 5. Analyze Slice and dice data-sets to understand odd behavior Use machine learning to detect patterns invisible for humans Convert data into insights 6. Predict Predictive models 7. Act Visualization layer Mobile applications Tailored to specific situation

8 Industry Case Studies Protecting Business Assets & High Value Commodities with IoT and Analytics IoT in Process Performance, Effiency & Predictive Maintenance

9 Protecting Business Assets & High Value Commodities with IoT and Analytics

10 People Data from a Different Perspective Questionable People Behaviour Movement outside of the norm Movement around High Risk Incidents Irregular Sequences and Links of Behaviour & Movement Network Analytics Who s talking to whom and how does it relate to Behaviours above Joining Intel from External Sources & Social Media Manipulation of Security Systems & Procedures

11 Joining People Data with Plant & Other Processes Process Performance Anomalies Performance vs. Expected Target Unplanned Stoppages/Breakdowns & People at the Brim of this Irregular Sequences and Links of Behaviour & Movement Irregular Behaviour of Systems Failure & manipulation of Monitoring & Security Controls Missing Data & Audits of Data Sequencing

12 Bringing it all together Adding Product Losses to the Mix Manipulation of Checks & Balances Anomalies in Product Recoveries Irregular behaviour in areas with open access to High Value Products External Sources of information Text Analytics on Unstructured information Reported Social Media Behaviour monitoring

13 Successes Thus Far Successful Prediction & Foiling of Theft to the Value of several Million $$$ Involved Clustering, Trees Methods, Nonlinear Modeling & Sequence & Link Analysis Monitoring & Controls (Security Assurance) Reduction in High Risk Incidents by effective audits, mapping of gaps identified Live Risk Management by Predictive/Prescriptive Analytics Analytical/Data links from Source to Sales Ability to Quantify Actual losses, based on Predictive analytics

14 Keys to Success Functional Task Team including: Domain/Business/Process Experts Security/Intelligence Industry Leading Expert Analytics Expertise & Extensive Experience Technology Expertise Correct & Efficient Deployment & Training Management Buy-in & Support

15 The Top-Drive Over 2000 sensors Mandatory total service every 5 year VERY expensive Designed to run at 100% for First analyses shows that critical parts don t runs that much A top drive is a mechanical device on a drilling rig that provides clockwise torque to the drill string to drill a borehole. It is an alternative to the rotary table and kelly drive. It is located at the swivel's place below the traveling block and moves vertically up and down the derrick.

16 Automatic detection of corrosion If we can t measure it We can t trace it

17 The Case.

18 Platform Inspections The Old Way

19 Platform Inspection the new way

20 Next Generation 1.0 Manually operated drone + Machine Learning for auto detection of corrosion

21 Next Generation 2.0 Fully automated drone inspection + Machine Learning for corrosion detection and tracking All flights recorded, and can be done again automatic autonomous Drone and inspection

22 The Red Challenge Also need to add shape and surface structure

23 Where to start? Convert this To this, pixel by pixel

24 Some examples

25 Predictive Maintenance of the Drone The picture could show sensor data from a case were the drone crashed because of a rotor failure. Available data was analysed and showed that the cause showed up in the data several flight hours before the crash. If we add this intelligence to the drone, the drone will take a decision to fly back to base and ask for the rotor to be replaced.

26 Process optimization

27 Industrial Dashboards with Alarms

28 Architecture Overview Where does it fit?

29 The Hawk use case

30 Connected Intelligent Drones Drones Mini computer Sensors Precise Prediction IoT Cloud Machine Learning All presented in a live Dashboard

31 IoT Intelligent Drones Intelligent Drones with sensors and Edge ML Our Machine Learning for Drones will make the drone intelligent Examples of sensors - Sound - Light - GPS - Dust - Gas - Vibration - Gyro - Air Quality / Air Pollution (many different sensors)

32 Q&A Thank You & Please come and see us at our Stand