Air Transport Industry-Cognitive Computing for Potential Information Mining

Case Background

In the era of globalization, there are tens of thousands of flights in the air every day. Although the incidence of aircraft accidents is extremely low, if it happens unfortunately, it will often lead to major disasters. Therefore, aircraft maintenance is an important step to ensure the safety of aircraft flight. A certain airline has accumulated many aircraft maintenance records for a long time, most of them are unstructured data and there is no uniform standard format for the report. However, when the aircraft fails, they always rely on the long-term accumulated maintenance experience to find out the causes of the accidents, which makes these huge data ineffective.


Import IBM Watson and obtain structured and unstructured data from an airline's ERP system environment, including technical manuals, non-routine records, technician notes, inventory records, troubleshooting records, material cost data and history of flight accidents. After loading the data, natural language processing and advanced content analysis techniques are used to transform the original unstructured data into potential factors for accidents. For example, it is found that the frequency of tire replacement demand increases due to high temperature and large passenger flow in summer, so that potential risk can be found and tire can be replaced early to avoid accidents.

In the past, technicians mostly relied on "past experience" to solve various problems. Now, through Watson's Cognitive Computing, enterprises can make up for the lack of human intuition, absorb larger and more diverse data, sift through years of unstructured historical data, find the root causes of accidents and direct the right solutions.

In addition, if problems occur in flight, the crew can immediately report them to the ground operators, and Watson will search for relevant data on similar incidents in the past (necessary materials for the aircraft, maintenance time, etc.) and compare the data to find a solution; when maintenance technicians carry out maintenance, they can input the adopted solution into Watson system and continuously enrich the internal knowledge base of Watson. Through Watson, maintenance technicians can also identify the types and trends of accidents in different seasons and provide these findings to equipment manufacturers for improvement.


  • The maintenance lead time is shorten by 90%, and the efficiency is improved by nearly 10 times
  • The system greatly simplifies technicians' search for and analysis of data to determine the cause of the incident, allowing them to spend less time on maintenance and more time on air travel.

  • Reduce flight delays and cancellations
  • Airlines reduced the number of delays and cancellations due to maintenance problems, greatly enhancing customer satisfaction and operation efficiency.

  • Simplification decision-making
  • Cognitive solutions relieve technicians' burden of finding answers to complex accident questions and reduce the time they spend searching, exporting and analyzing data, enabling them to quickly diagnose and solve problems.


Relevant Cases:

  1. Energy Industry-Data Integration and Decision Assistance
  2. InSites Consulting Corp. Establishes the Sentiment Dictionary Based on IBM SPSS Modeler
  3. Direct Wine Conducts Differentiated Marketing Based on IBM SPSS Modeler

AsiaAnalytics Taiwan Ltd.

5F, No. 356, Sec. 1, Neihu Rd., Neihu Dist., Taipei City 11493, Taiwan (R.O.C.)

TEL:+886 2 7728 7958 FAX: +886 2 2627 0667