ibm predictive analytics iot presentation

34
Predictive Analytics for IoT Michael Adendorff Architect, STSM IBM IBM Predictive Maintenance and Quality [email protected]

Upload: ianskerrett

Post on 10-May-2015

772 views

Category:

Software


5 download

TRANSCRIPT

Page 1: IBM Predictive analytics IoT Presentation

Predictive Analytics for IoT

Michael Adendorff Architect, STSM IBM

IBM Predictive Maintenance and Quality [email protected]

Page 2: IBM Predictive analytics IoT Presentation
Page 3: IBM Predictive analytics IoT Presentation

Evidence , Clues > Failure Prediction

Page 4: IBM Predictive analytics IoT Presentation

Predictive Analytics

Valuable Insight

Page 5: IBM Predictive analytics IoT Presentation

Maintenance Insight

Page 6: IBM Predictive analytics IoT Presentation

Maintenance Insight

Failure Risk: Under Maintained Equipment

Page 7: IBM Predictive analytics IoT Presentation

Maintenance Insight

Wasted $$$$$: Over maintained equipment

Page 8: IBM Predictive analytics IoT Presentation

Work Order

Urgent Inspection Required: High probability of failure

High risk of failure before next scheduled maintenance

Page 9: IBM Predictive analytics IoT Presentation

Maintenance Schedule Update Request

Bring forward scheduled maintenance to Jul 5

Bring forward scheduled maintenance to Aug 7

Delay scheduled maintenance to Dec 15

Page 10: IBM Predictive analytics IoT Presentation

Parts Requirements Forecast : Main Bearing

June: July: Aug:

10 3 22

9 20 12

7 21 14

12 15 17

Page 11: IBM Predictive analytics IoT Presentation

Business Results : Predictive Maintenance

Downtime

Unplanned

Planned

Page 12: IBM Predictive analytics IoT Presentation

Predictive Analytics

Valuable Insight

How does it work?

Page 13: IBM Predictive analytics IoT Presentation

Simplistic Illustration

Historic Data

Failure Records

Vibration Levels

Correlation

Failu

re C

ou

nt

Vibration Level

More failures have been witnessed when vibration levels are high

Page 14: IBM Predictive analytics IoT Presentation

Univariate Model

Failu

re C

ou

nt

Vibration Level

Vibration Level

P Failure Confidence

< 0.1 0.1 % 2%

0.-0.5 1% 3%

0.5 – 2 3% 5%

2 – 5 15% 10%

+5 98% 80%

p(fail)

Simple univariate models are generally not very accurate. This one looks better than it is. High vibration strongly correlated with failure as it is a lagging indicator. Need leading indicators to predict.

Page 15: IBM Predictive analytics IoT Presentation

Multivariate model

p(fail)

More accurate than the univariate model, but raw input data never reveals the whole story.

Correlates failures with combinations between multiple input variables

Historic Data

Page 16: IBM Predictive analytics IoT Presentation

Advanced Data Prep + Ensemble Models

More accurate than the univariate model, but raw input data never reveals the whole story.

Historic Data

p(fail)

E(fail date)

Page 17: IBM Predictive analytics IoT Presentation

Advanced Data Prep + Ensemble Models

More accurate than the univariate model, but raw input data never reveals the whole story.

Historic Data

Cumulative Cycles = f(speed, operating hours)

p(fail)

E(fail date)

Page 18: IBM Predictive analytics IoT Presentation

Advanced Data Prep + Ensemble Models

More accurate than the univariate model, but raw input data never reveals the whole story.

Historic Data

Cumulative Fatigue Load = f(Cycles, Speed)

p(fail)

E(fail date)

Page 19: IBM Predictive analytics IoT Presentation

Advanced Data Prep + Ensemble Models

More accurate than the univariate model, but raw input data never reveals the whole story.

Historic Data

Wear Damage Forecast

p(fail)

E(fail date)

Page 20: IBM Predictive analytics IoT Presentation

Advanced Data Prep + Ensemble Models

p(fail)

More accurate than the univariate model, but raw input data never reveals the whole story.

Historic Data

Wear Damage Forecast

E(fail date)

Wear Modeling

Page 21: IBM Predictive analytics IoT Presentation

Advanced Data Prep + Ensemble Models

More accurate than the univariate model, but raw input data never reveals the whole story.

Historic Data

Fatigue Damage Forecast

p(fail)

E(fail date)

Page 22: IBM Predictive analytics IoT Presentation

Advanced Data Prep + Ensemble Models

p(fail)

More accurate than the univariate model, but raw input data never reveals the whole story.

Historic Data

Wear Damage Forecast

E(fail date)

Fatigue Modeling

Page 23: IBM Predictive analytics IoT Presentation

Advanced Data Prep + Ensemble Models

Building models like this requires brute force number crunching as well as skills and knowledge. Payoff comes from more accurate predictions – but – it doesn’t end here.

Historic Data

Time series forecast + Combination Model

p(fail)

E(fail date)

Page 24: IBM Predictive analytics IoT Presentation

Advanced Data Prep + Ensemble Models

Historic Data

Expected failure date is more actionable than current probability of failure

Building models like this requires brute force number crunching as well as skills and knowledge. Payoff comes from more accurate predictions – but – it doesn’t end here.

p(fail)

E(fail date)

Page 25: IBM Predictive analytics IoT Presentation

Advanced Data Prep + Ensemble Models

Historic Data

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut lacinia semper gravida. Morbi vel orci in leo malesuada malesuada in ac enim. Nam pulvinar nec enim in venenatis. In nibh turpis, sodales at fermentum in

Sensors don’t record every causal factor. Text analytics is used to fill in some of the blanks.

p(fail)

E(fail date)

Page 26: IBM Predictive analytics IoT Presentation

Predictive Analytics

Valuable Insight

Building models is only half the fun. Next step – OPERATIONALIZE

Page 27: IBM Predictive analytics IoT Presentation

Feed Data

APIs for: • Describing target data structures • Describing calculations and aggregations • Running analytics • Exposing analytic results

REST Historian DB

Web

Se

rvic

e

MQTT Other

Page 28: IBM Predictive analytics IoT Presentation

Data flows into DB in realtime

Event

Master Data

Profile

KPI

Page 29: IBM Predictive analytics IoT Presentation

Predictive Analytics done in realtime

Event

Master Data

Profile

KPI

p(fail)

E(fail date)

Page 30: IBM Predictive analytics IoT Presentation

Predictive Analytics done in realtime

Event

Master Data

Profile

KPI

p(fail)

E(fail date)

Predictive Outputs fed back as new events

Page 31: IBM Predictive analytics IoT Presentation

Deciding on Recommended Actions

Event

Profile Action

KPI

Page 32: IBM Predictive analytics IoT Presentation

Taking Action

REST DB

Web

Ser

vice

FTP Other

Page 33: IBM Predictive analytics IoT Presentation

Valuable Insight

Build Models 1) Assemble historic data 2) Attempt to correlate historical data with a

known target 3) Improve results by putting more thought

about preparing inputs and algorithm selection

Operationalize 1) Feed raw data 2) Describe calculation and aggregation 3) Perform analytics 4) Carry out decision logic 5) Feed results 6) Retrain models regularly

Page 34: IBM Predictive analytics IoT Presentation

Questions?

Michael Adendorff Architect, STSM IBM

IBM Predictive Maintenance and Quality [email protected]