Save lead time to recognize & provide an exactly matching concession in a database of concessions by utilizing the non-conformity images, non-conformity type, impacted drawing number & location of the impact.

Solution

• By using data analytics concepts, the concessions are bucketed into different categories like type of non-conformity & impacted drawing etc., for each perimeter and for each program under AXISCADES’ responsibility.
• The data obtained in step 1 & the image of non-conformity highlighted on ZAMIZ drawing & actual aircraft will be taken into AIML model.
• The AIML model will be based on Artificial Neural Network or Case Based Reasoning Framework.

Business Value:

Productivity: AI / ML based application will identify relevant or very similar concessions very quickly, thereby improving the productivity.

Accuracy: Reduces time taken for each concession search from 15 minutes - 1 hour to less than a minute without compromising the data integrity.

Efficiency: The highly efficient, accurate algorithm will be chosen out of existing various AI/ML based algorithms, which guarantees the improvement in efficiency.

AIML application to recognize & provide an exactly matching repair solution from the database by reading damage images, explanation, impacted drawing number & location of the impact.

Solution

• Model uses latest trends in image, text processing and scanning to read the damages.
• Data analysed using ML algorithms and classify the damage type, read the damage details, search for similar type of damage / damages and propose the repair solutions by referring to SRM, by engineering judgment, using tools & techniques.
• Once the model is trained with historical data, the repair solution can be generated instantly for common type of repairs. The uncommon repairs will also be fed back in the model once the solution is generated manually.

Business Value:

Efficiency: Standardized airliner damage reports will reduce communication loops & hence less number of TRM lines will be necessary to close a repair case.

Quicker resolutions: ML can provide repair solution/proposals for repetitive tasks quickly and less lead time to address AoG cases. Harmonized solutions will reduce check, approval & implementation time.

Accuracy: The repair solutions will be based on the historic data, hence approved methodologies.

Extract information (data drivers) from relevant sheet metal, machining, assembly, tubes & pipes drawings using automation.

Solution

• Relevant parameters from the drawings which are in PDF / image format are extracted by using Artificial Intelligence powered visual comprehension.
• Relevant parameters from the drawings which are in other documents (BoM, material norms, etc.) is also digitized using AI and then cross-referenced automatically.
• Engineer focuses on QC of AI powered entries allowing simultaneous data entry and QC.