Big Data and Analytics Creating Actionable Intelligence

Transcription

1 Big Data and Analytics Creating Actionable Intelligence Wednesday October 17, :30 pm - 4:30 pm Shane Benfield Boeing Global Services Analytics

2 Key Themes for this session Focus on outcomes before data Collaboration is key Analytics programs are scalable Connect with us socially #NBAA18 2

3 Analytics Projects The Promise Today s Reality Key Challenges Big Data Analytics Tools & Platform Big Value Only 25% Satisfied Highly satisfied, 6% Satisfied 19% Neutral 54% Highly dissatisfied, 3% Dissatisfied 18% Tools Culture Data Access Talent / Expertise Other 4% 16% 27% 27% 26% How satisfied are you with results of current analytics projects? What are your key challenges to achieving improved results from analytics? Source: Boeing Global Services Customer Conference (11/2017) 3

4 Our Perspective Customer Value Through an Analytics Ecosystem Provide Customer Choice Digital Solutions = Real-Time Applications + Integration & Automation + Prognostics & Optimization Fully outsourced 3% Completely on your own 7% + Analytics Projects Insights = Big Data Data + Analytics Tools & Platform + Aerospace Engineering Insights Digital solutions with embeded analytics 63% Third-party tools/data/ expertise 27% How do you plan to maximize the value of analytics? Source: Boeing Global Services Customer Conference (11/2017) 4

5 End-to-End Analytics Customer Value Self-Service Analytics/BI Analytics Consulting and Support Data / Content Domain expertise Data scientists OEM insights Data and Analytics Platform Analytics Enabled Digital Solutions Data Periodic Projects Continuous Optimization 5

6 Two Questions Connect with us socially #NBAA18 6

7 Question 1: Hype Not Hype? Big Data, Data Mining, Data Analytics, Artificial Intelligence A. It is marketing hype, vaporware, an empty promise B. A lot of hype, but I do see a core that is real and potentially promising C. Some hype, and my company does do some of these activities and we have had some success D. My company uses it everyday to help our operations E. I have no idea what these terms mean Connect with us socially #NBAA18 7

8 Question 2: To what extent is your operation using data and data analytics? A. We have our own in-house analytics team B. We use outside consultants to handle our analytics needs C. We use analytics enabled applications in our operations D. We use consultants and analytics applications E. To my knowledge we don t use any analytics Connect with us socially #NBAA18 8

9 What does a BA operator care about? Safety Customer Service Demand Weather, ATC or other impacts to day-of-operation Operating Costs Maintenance Events Fuel Management Special Events Connect with us socially #NBAA18 9

10 Aerospace data generation growing exponentially Digital transformation of aerospace is driving the creation of real-time data growth Petabytes Past Present Future Terabytes Gigabytes Megabytes Time Aircraft and Operational data from increased onboard sensors and digitalized operations Connect with us socially #NBAA18 10

11 What do we mean by data? Direct data sensors readings Derived data calculated from sensors or other onboard information Operational data all non-aircraft data from flight ops to maintenance OEM and Industry data - like ATC, weather, and information provided by the manufacturers and suppliers Connect with us socially #NBAA18 11

12 What is data analytics? Transforming data into insights Used for revealing opportunities Method to improve decision making Tools to optimize processes Connect with us socially #NBAA18 12

13 A brief overview of the four analytics types What will happen? How can we make it happen? Prescriptive Analytics Why did it happen? Predictive Analytics What happened? Diagnostic Analytics Descriptive Analytics Connect with us socially #NBAA18 13

14 It is more than statistics, it is a rich set of tools and methods Examples: Modeling & Simulation Statistics Expert Systems Classification Engines Machine Learning Network Analysis Artificial Intelligence Semantic Search Text Analytics Non-Linear Optimization Neural Networks Mathematics Data Visualization Pattern Recognition Deep Learning Indexing Connect with us socially #NBAA18 14

15 In aviation technical operations, data and analytics represent physical objects and/or actions; real-world stuff. For example Connect with us socially #NBAA18 15

16 One method used to create a predictive alert It takes historical data, data analytics, and domain expertise Develop Prediction Test & Verify Validate in Ops Statistical Analysis Release for Use Sensor inputs OEM Knowledge Connect with us socially #NBAA18 16

17 Another Question Connect with us socially #NBAA18 17

18 Question 3: What are your key challenges to achieving improved results from analytics? A. Lack of access or issues with data analytics tools B. Acceptance by, or implementation within, the operational culture C. Access to data either internal or external to the organization D. Access to analytics talent and/or expertise Connect with us socially #NBAA18 18

19 Approach to maximize value Have clarity on business goals Optimize the system Identify the technology - applications Prepare the people and processes Connect with us socially #NBAA18 19

20 It all starts with goals Understand the desired business outcome or goals before going after methods, technology, or data. Efficiency Performance Safety Economy Schedule Reliability Connect with us socially #NBAA18 20

21 Then we work back from the desired results with goals What actions will be needed? What decisions drive these actions? What situations require these decisions? Connect with us socially #NBAA18 21

22 This drives the right mix of analytics and data This method also supports addressing the people and process issues at a system level Connect with us socially #NBAA18 22

23 Last Question Connect with us socially #NBAA18 23

24 Question 4: Of the issues listed below, which one are you most concerned about? A. Increasing Revenue B. Fuel Spend C. Improving Fleet Reliability D. Operational Costs E. Competitive Pressure Connect with us socially #NBAA18 24

25 Example Use Cases Connect with us socially #NBAA18 25

26 Analytics applied to Flight Optimization One major saved 350 min/day on 4K flights and reduced fuel by 14M lbs./year Alerting on opportunities for optimization not available in a filed flight plan Continuously looking for postdeparture wind-optimal shortcuts (10%- 20%) Monitors surrounding traffic, up-todate weather, and airspace constraints Connect with us socially #NBAA18 26

27 Predictive Maintenance Eliminating unscheduled maintenance events reduces delays or cancellations Predictive alert on 777 Integrated Drive Generator avoided an in-service delay or cancellation and saved up to $300,000 in repair costs. Connect with us socially #NBAA18 27

28 Fuel analysis saves GOL 20 million kilos of fuel Fuel savings of 1% - 3% have been achieved for some operators GOL Transportes Aereos using fuel data analytics Identify potential savings and track savings across all phases of flight Benchmark performance and optimize flight plans Connect with us socially #NBAA18 28

29 C-17 Lifecycle Optimization Program New analytics applications to increase value and readiness In-flight and historical data combined with maintenance plans and other data helped increase readiness and lower lifecycle costs Fleet performance trending and analysis Fuel optimization Part failure prevention Improved maintenance operations Connect with us socially #NBAA18 29

30 Key Learnings Data Analytics boosts competitiveness Focus on achieving a business outcome Data + statistics + tools + domain expertise = effective analytics solutions Analytics programs are scalable Connect with us socially #NBAA18 30

31