AI-POWERED PREDICTIVE MAINTENANCE
© 2018 Findability Sciences
❑ Unifies Structured, Unstructured,
Internal and External Data.
❑ Text Analytics, Natural Language
Processing, Cognitive Analytics.
❑ Artificial Intelligence (AI) provided
through self-learning, fully automatic
multi-modelling algorithm.
❑ Cloud Based. SaaS or DaaS/PRaSS
FINDABILITY PLATFORM®
❑ Financial Services
❑ Manufacturing
❑ Pharma/Health
❑ Retail
❑ General Trading
❑ Higher Education
❑ Automotive/Transportation
INDUSTRIES SERVED
❑ Digital Body Language (DBL)
❑ Customer Persona
❑ People Analytics
❑ Fraud Detections
❑ Price Predictions
❑ Inventory Predictions
❑ Predictive Maintenance
❑ ChatBot
SOLUTIONS
FINDABILITY SCIENCES ➢ Developer of award winning Findability Platform®, a leading
Bigdata, Cognitive and AI Platform.
➢ SoftBank Corp’s Portfolio Company.
➢ Joint Venture with SoftBank in Japan.
➢ Erstwhile Member of Partner Advisory Board of IBM Watson™.
➢ Investor in Boston based Soft10 Inc.
➢ Global Presence: Boston, Tokyo, Mumbai and Aurangabad.
ECOSYSTEM PARTNERS
© 2018 Findability Sciences
FINDABILITY SCIENCES KK
• Strong presence in Japan
• ~1,000 B2B sales staff
• Telecom, AI, Robotics
• Watson Partnership
• Largest vision fund
• Tech IP
• Service delivery
• Industry agnostic solution
• Watson partnership and
experiences
AUTOMATED PREDICTION, AI CHAT-BOT AND NLP Q&A, COGNITIVE & BIG DATA
ANALYTICS © 2018 Findability Sciences. Confidential
AI-POWERED PREDICTIVE MAINTENANCE FOR AIRLINE
FINDABILITY PLATFORM®
CUSTOMER
International airline in United States
PROBLEM
Wanted to predict defects specific to Air Conditioner ( ATA
21), Pressurization (ATA 36), and flight control (ATA 27)
failures in fleets.
SOLUTION
Findability Platform® self learning AI machine to identify
component failure in advance.
• Data:Defects and delays data related to A320 and E190
fleet specific to ATA 21, ATA 36 and ATA 27
• Training: Automatically learns from historical data
• Prediction: Predict for next possible component failure
BENEFITS
• 88% accurate prediction
• Minimize corrective and breakdown maintenance
• Help to maintain satisfactory equipment conditions, and
improves fleet reliability
• Minimizes the risk of aircraft failure
• Saves significant money
© 2018 Findability Sciences
IDENTIFY COMPONENT FAILURE
CUSTOMER
International airline in western Pacific region
PROBLEM
Wanted to know flight engine failure in advance to convert
unscheduled maintenance to planned maintenance and
remove chaos
SOLUTION
Findability Platform® self learning AI machine to predict
failure from 60-90 days in advance.
• Data: Flight engine log, flight information system log,
pollution data, flight path data
• Training: Automatically learns from failure engines (left,
right, international, domestic).
• Prediction: Predict for next possible engines failure.
BENEFITS
• 90% accurate prediction.
• 97% accuracy in top 3 decile. Customer uses top 3 decile
result to schedule immediate engine inspection.
© 2018 Findability Sciences
ENGINE FAILURE PREDICTION
PREDICTION APPROACH - FINDABILITY SCIENCES
HISTORICAL INFORMATION
insight
SELF- LEARNING
Predict
PREDICTIVE ANALYSIS
© 2018 Findability Sciences
Cleaning, Combining & enrichment of
data
With out human intervention the
solution develops predictive models
Continuous improvement and
automation of the result
With the use of API connection the
solution automatically runs and
complies heterogeneous data
Data is processed through a self-learning
algorithm to develop models depending
on use case
The solution automatically calculates
relevant score and ranks models in order
of probability
PREDICTIVE MAINTENANCE APPROACH - FINDABILITY SCIENCES
FAILURE HISTORY - The failure history of a machine or
component within the machine.
MACHINE CONDITION - The operating characteristics
of a machine, e.g. data collected from sensors.
REPAIR HISTORY - The repair history of a machine, e.g.
previous maintenance records, components replaced,
maintenance activities performed.
MACHINE FEATURES - The features of machine or
components, e.g. production date, technical
specifications
OPERATING CONDITIONS - Environmental features
that may influence a machine’s performance, e.g.
location, temperature, other interactions.
SELF-LEARNING - Learns from historical data to
generate continuously updated results that accommodate
ever-changing data
AUTOMATIC - Unsupervised, fully autonomous multi-
step modelling and prediction process that produces
results faster
NON PARAMETRIC - Based on Conditional Probabilities
no assumptions to limit the best models.
MULTI MODELING - Generates and selects multiple
models from a data pattern, making use of vast volumes
of unstructured data in addition to structured data
Detect anomalies in equipment or system performance or
functionality.
Predict whether an asset may fail in the near future.
Estimate the remaining useful life of an asset.
Identify the main causes of failure of an asset.
Identify what maintenance actions need to be done, by
when, on an asset
Collect & Clean Data
HISTORICAL INFORMATION
insight
Identify Pattern
MACHINE LEARNING
Make Prediction
Predict
PREDICTIVE ANALYSIS
© 2018 Findability Sciences
© 2018 Findability Sciences
PREDICTION PLATFORM METHODOLOGY
SELF LEARNING, AUTO MULTI-MODELING,
FULLY AUTOMATIC PREDICTION
HISTORICAL AIRCRAFT
INFORMATION
ONGOING
DAILY/WEEKLY/ MONTHLY DATA
1
SELF-LEARNING PREDICTION
M1
M2
M3
M4
M..
Findability Platform®
AUTO MULTI-MODELLING
2
3 4
5
DAILY/ WEEKLY/
MONTHLY
PREDICTIONS WITH
ASSOCIATED
PROBABILITY
• Reduce operational risk of mission critical
equipment.
• Increase rate of return on assets by
predicting failures before they occur.
• Control cost of maintenance by enabling
just-in-time maintenance operations.
• Lower inventory costs by reducing inventory
levels by predicting the reorder point.
• Discover patterns connected to various
maintenance problems.
• Estimate remaining lifespan of assets.
• Recommend timely maintenance activities.
BENEFIT
© 2018 Findability Sciences
Build future with AI
Vivek Vij
vivek@findabilitysciences.com
Direct: +1 781-353-2466
Findability Sciences 300 TravelCenters Drive STE 4690 Woburn MA 01801

AI-powered predictive maintenance for Airlines

  • 1.
    AI-POWERED PREDICTIVE MAINTENANCE ©2018 Findability Sciences
  • 2.
    ❑ Unifies Structured,Unstructured, Internal and External Data. ❑ Text Analytics, Natural Language Processing, Cognitive Analytics. ❑ Artificial Intelligence (AI) provided through self-learning, fully automatic multi-modelling algorithm. ❑ Cloud Based. SaaS or DaaS/PRaSS FINDABILITY PLATFORM® ❑ Financial Services ❑ Manufacturing ❑ Pharma/Health ❑ Retail ❑ General Trading ❑ Higher Education ❑ Automotive/Transportation INDUSTRIES SERVED ❑ Digital Body Language (DBL) ❑ Customer Persona ❑ People Analytics ❑ Fraud Detections ❑ Price Predictions ❑ Inventory Predictions ❑ Predictive Maintenance ❑ ChatBot SOLUTIONS FINDABILITY SCIENCES ➢ Developer of award winning Findability Platform®, a leading Bigdata, Cognitive and AI Platform. ➢ SoftBank Corp’s Portfolio Company. ➢ Joint Venture with SoftBank in Japan. ➢ Erstwhile Member of Partner Advisory Board of IBM Watson™. ➢ Investor in Boston based Soft10 Inc. ➢ Global Presence: Boston, Tokyo, Mumbai and Aurangabad. ECOSYSTEM PARTNERS © 2018 Findability Sciences
  • 3.
    FINDABILITY SCIENCES KK •Strong presence in Japan • ~1,000 B2B sales staff • Telecom, AI, Robotics • Watson Partnership • Largest vision fund • Tech IP • Service delivery • Industry agnostic solution • Watson partnership and experiences AUTOMATED PREDICTION, AI CHAT-BOT AND NLP Q&A, COGNITIVE & BIG DATA ANALYTICS © 2018 Findability Sciences. Confidential
  • 4.
    AI-POWERED PREDICTIVE MAINTENANCEFOR AIRLINE FINDABILITY PLATFORM®
  • 5.
    CUSTOMER International airline inUnited States PROBLEM Wanted to predict defects specific to Air Conditioner ( ATA 21), Pressurization (ATA 36), and flight control (ATA 27) failures in fleets. SOLUTION Findability Platform® self learning AI machine to identify component failure in advance. • Data:Defects and delays data related to A320 and E190 fleet specific to ATA 21, ATA 36 and ATA 27 • Training: Automatically learns from historical data • Prediction: Predict for next possible component failure BENEFITS • 88% accurate prediction • Minimize corrective and breakdown maintenance • Help to maintain satisfactory equipment conditions, and improves fleet reliability • Minimizes the risk of aircraft failure • Saves significant money © 2018 Findability Sciences IDENTIFY COMPONENT FAILURE
  • 6.
    CUSTOMER International airline inwestern Pacific region PROBLEM Wanted to know flight engine failure in advance to convert unscheduled maintenance to planned maintenance and remove chaos SOLUTION Findability Platform® self learning AI machine to predict failure from 60-90 days in advance. • Data: Flight engine log, flight information system log, pollution data, flight path data • Training: Automatically learns from failure engines (left, right, international, domestic). • Prediction: Predict for next possible engines failure. BENEFITS • 90% accurate prediction. • 97% accuracy in top 3 decile. Customer uses top 3 decile result to schedule immediate engine inspection. © 2018 Findability Sciences ENGINE FAILURE PREDICTION
  • 7.
    PREDICTION APPROACH -FINDABILITY SCIENCES HISTORICAL INFORMATION insight SELF- LEARNING Predict PREDICTIVE ANALYSIS © 2018 Findability Sciences Cleaning, Combining & enrichment of data With out human intervention the solution develops predictive models Continuous improvement and automation of the result With the use of API connection the solution automatically runs and complies heterogeneous data Data is processed through a self-learning algorithm to develop models depending on use case The solution automatically calculates relevant score and ranks models in order of probability
  • 8.
    PREDICTIVE MAINTENANCE APPROACH- FINDABILITY SCIENCES FAILURE HISTORY - The failure history of a machine or component within the machine. MACHINE CONDITION - The operating characteristics of a machine, e.g. data collected from sensors. REPAIR HISTORY - The repair history of a machine, e.g. previous maintenance records, components replaced, maintenance activities performed. MACHINE FEATURES - The features of machine or components, e.g. production date, technical specifications OPERATING CONDITIONS - Environmental features that may influence a machine’s performance, e.g. location, temperature, other interactions. SELF-LEARNING - Learns from historical data to generate continuously updated results that accommodate ever-changing data AUTOMATIC - Unsupervised, fully autonomous multi- step modelling and prediction process that produces results faster NON PARAMETRIC - Based on Conditional Probabilities no assumptions to limit the best models. MULTI MODELING - Generates and selects multiple models from a data pattern, making use of vast volumes of unstructured data in addition to structured data Detect anomalies in equipment or system performance or functionality. Predict whether an asset may fail in the near future. Estimate the remaining useful life of an asset. Identify the main causes of failure of an asset. Identify what maintenance actions need to be done, by when, on an asset Collect & Clean Data HISTORICAL INFORMATION insight Identify Pattern MACHINE LEARNING Make Prediction Predict PREDICTIVE ANALYSIS © 2018 Findability Sciences
  • 9.
    © 2018 FindabilitySciences PREDICTION PLATFORM METHODOLOGY SELF LEARNING, AUTO MULTI-MODELING, FULLY AUTOMATIC PREDICTION HISTORICAL AIRCRAFT INFORMATION ONGOING DAILY/WEEKLY/ MONTHLY DATA 1 SELF-LEARNING PREDICTION M1 M2 M3 M4 M.. Findability Platform® AUTO MULTI-MODELLING 2 3 4 5 DAILY/ WEEKLY/ MONTHLY PREDICTIONS WITH ASSOCIATED PROBABILITY
  • 10.
    • Reduce operationalrisk of mission critical equipment. • Increase rate of return on assets by predicting failures before they occur. • Control cost of maintenance by enabling just-in-time maintenance operations. • Lower inventory costs by reducing inventory levels by predicting the reorder point. • Discover patterns connected to various maintenance problems. • Estimate remaining lifespan of assets. • Recommend timely maintenance activities. BENEFIT © 2018 Findability Sciences
  • 11.
    Build future withAI Vivek Vij vivek@findabilitysciences.com Direct: +1 781-353-2466 Findability Sciences 300 TravelCenters Drive STE 4690 Woburn MA 01801

Editor's Notes

  • #8 Unprecedented amount of aircraft, airline-aviation operational data has opened up the potential for doing predictive maintenance – the capability to spot an emerging issue before it may impact schedule operations.  With progress in sensor technology and data processing techniques, structural health monitoring (SHM) can lead the possibility of planning maintenance based on an actual state of aircraft rather than on a fixed schedule
  • #9 Unprecedented amount of aircraft, airline-aviation operational data has opened up the potential for doing predictive maintenance – the capability to spot an emerging issue before it may impact schedule operations.  With progress in sensor technology and data processing techniques, structural health monitoring (SHM) can lead the possibility of planning maintenance based on an actual state of aircraft rather than on a fixed schedule