www.aero-defence.tech
16
'26
Written on Modified on
Pratt & Whitney advances engine inspections with AI-powered technology
This technology enables a step change in how inspections are performed, enhancing consistency and efficiency across global MRO operations for commercial, civil and military engines.
www.rtx.com
Pratt & Whitney boosts engine inspection capabilities by acquiring Amsterdam-based Aiir innovations and integrating its AI-assisted borescope software. Credit: Pratt & Whitney
Pratt & Whitney, an RTX business, has expanded its maintenance framework through the acquisition and structural integration of Amsterdam-based Aiir Innovations. The acquisition introduces specialized artificial intelligence (AI) borescope inspection software into the engine manufacturer's global maintenance, repair, and overhaul (MRO) facilities, automating key analysis tasks for commercial, civil, and military aviation powerplants.
Borescope Video Analytics and Processing Efficiency
The core software technology applies computer vision and deep learning models directly to raw borescope video streams recorded during internal engine inspections. Instead of relying exclusively on manual optical frame reviews by technicians, the algorithmic layer detects, highlights, and catalogs anomalies in real time, delivering faster and more repeatable engine condition assessments.
The system has already been deployed to commercial airlines and third-party MRO providers, reducing overall inspection turnaround times. Pratt & Whitney has integrated the automated workflow into standard maintenance procedures for the V2500 engine family. Concurrently, the aviation OEM recently finalized validation pilot programs on its Geared Turbofan (GTF) commercial engines and the military F135 propulsion system, with formal deployment strategies underway to expand the software's footprint across its entire production and servicing infrastructure.
Feedback-Driven Learning Loops and Reporting Architecture
The platform utilizes an interactive feedback architecture to support long-term accuracy. As human inspectors review, confirm, adjust, or correct the automated defect classifications, the machine learning models adjust their internal weight parameters. This iterative learning loop aligns the software's predictive neural pathways with localized real-world inspection expertise over consecutive maintenance lifecycles.
Additional Context
This section details technical specifications not included in the original news release.
Aircraft gas turbines require routine Borescope Inspections (BSIs) to evaluate the structural integrity of internal hot-section components, such as high-pressure turbine blades, compressor vanes, and combustion liners, without executing a complete engine teardown. Inspectors thread a flexible fiber-optic insertion tube equipped with a micro-CMOS image sensor through narrow access ports, navigating complex internal spaces to visually check for material distress, micro-cracks, thermal coating degradation, and foreign object damage (FOD).
The Aiir Innovations software automates this visual mapping process by routing the live digital video stream through a Convolutional Neural Network (CNN) trained on thousands of labeled engine defect profiles. The system flags localized pixel patterns that signify structural deviations, while an integrated frame-stitching algorithm tracks individual blade counts during turbine rotation to map each defect to its exact spatial location.
The software then references the targeted internal component against digitized Aircraft Maintenance Manuals (AMMs) and Engine Manuals (EMs) via an automated look-up script. This utility allows the inspection system to automatically compare identified physical cracks or spalling dimensions against the engine's strict serviceability limits, streamlining the decision-making process for engineering teams.
Edited by Romila DSilva, Induportals Editor, with AI assistance.
Pratt & Whitney, an RTX business, has expanded its maintenance framework through the acquisition and structural integration of Amsterdam-based Aiir Innovations. The acquisition introduces specialized artificial intelligence (AI) borescope inspection software into the engine manufacturer's global maintenance, repair, and overhaul (MRO) facilities, automating key analysis tasks for commercial, civil, and military aviation powerplants.
Borescope Video Analytics and Processing Efficiency
The core software technology applies computer vision and deep learning models directly to raw borescope video streams recorded during internal engine inspections. Instead of relying exclusively on manual optical frame reviews by technicians, the algorithmic layer detects, highlights, and catalogs anomalies in real time, delivering faster and more repeatable engine condition assessments.
The system has already been deployed to commercial airlines and third-party MRO providers, reducing overall inspection turnaround times. Pratt & Whitney has integrated the automated workflow into standard maintenance procedures for the V2500 engine family. Concurrently, the aviation OEM recently finalized validation pilot programs on its Geared Turbofan (GTF) commercial engines and the military F135 propulsion system, with formal deployment strategies underway to expand the software's footprint across its entire production and servicing infrastructure.
Feedback-Driven Learning Loops and Reporting Architecture
The platform utilizes an interactive feedback architecture to support long-term accuracy. As human inspectors review, confirm, adjust, or correct the automated defect classifications, the machine learning models adjust their internal weight parameters. This iterative learning loop aligns the software's predictive neural pathways with localized real-world inspection expertise over consecutive maintenance lifecycles.
Additional Context
This section details technical specifications not included in the original news release.
Aircraft gas turbines require routine Borescope Inspections (BSIs) to evaluate the structural integrity of internal hot-section components, such as high-pressure turbine blades, compressor vanes, and combustion liners, without executing a complete engine teardown. Inspectors thread a flexible fiber-optic insertion tube equipped with a micro-CMOS image sensor through narrow access ports, navigating complex internal spaces to visually check for material distress, micro-cracks, thermal coating degradation, and foreign object damage (FOD).
The Aiir Innovations software automates this visual mapping process by routing the live digital video stream through a Convolutional Neural Network (CNN) trained on thousands of labeled engine defect profiles. The system flags localized pixel patterns that signify structural deviations, while an integrated frame-stitching algorithm tracks individual blade counts during turbine rotation to map each defect to its exact spatial location.
The software then references the targeted internal component against digitized Aircraft Maintenance Manuals (AMMs) and Engine Manuals (EMs) via an automated look-up script. This utility allows the inspection system to automatically compare identified physical cracks or spalling dimensions against the engine's strict serviceability limits, streamlining the decision-making process for engineering teams.
Edited by Romila DSilva, Induportals Editor, with AI assistance.
www.rtx.com
