FEATURE:AEROSPACE


engines on their mobile devices and receive real-time responses on parts availability. This helps to accelerate turnaround of leased engine assets. Instead of an inspector having to examine an engine and check part-by-part, Pratt & Whitney says Percept automates this inspection, and reduces time taken by nearly 90%. “The Percept tool helps reduce time and


effort involved in the pre-and-post lease analysis of aircraft engines,” says O Sung Kwon, Vice-president, Customer Support, Pratt & Whitney. “We have been working with Awiros, an Indian computer vision and AI start- up for the past few years to mature this technology; we are excited to be shifting from technology development to now bringing an operational product to the market.” And in January, Boeing filed a US


patent (US-20240029414-A1) for an AI-assisted inspection system for aircraft components “to assist a human operator in identifying anomalous areas of manufactured components, such as aircraft components.” Boeing’s application states that “by using a computational model to augment the human operator’s judgement and/or attention to detail, anomalies within the aircraft surfaces can be identified more accurately. In particular, the computational model is robust in part because it is trained with labelled images having varying degrees of detail, and because it uses images having varying degrees of detail to identify anomalies within the aircraft surfaces.” It goes on to say that the computational model could also be used to predict whether new anomalies will develop and how they would propagate across aircraft surfaces”. On 5 January, the importance of


accurate aircraft inspection was illustrated when a door plug blew out from Alaska Airlines flight 1282, a Boeing 737-9 MAX, leading to the rapid decompression of the plane’s cabin. An initial report by the National Transportation Safety Board (NTSB) found that as many as four bolts were likely missing from the emergency exit door panel.


Laying the groundwork Manufacturing of aerospace vehicle components calls for a full 360-degree view of all parts involved, along with real-time insights detailing areas that need focus. In this industry, it is vital that errors and waste are minimised from as early in the supply chain as possible. Along with machine vision and imaging


capabilities, all departments within an aerospace OEM need a fully democratised view of project progress, for effective collaboration. At Lockheed Martin UK, for example, a


paperless manufacturing execution system is underpinned by real-time imaging data generated by existing parts, captured by smart tablets. Operators can overlay CAD model data over objects, at the point of installation, helping to drive down the manufacturing learning curve, as well as costs and time. This is overseen by 205 industrial engineers – an estimated 40% of whom directly interact with computer vision technology carried out by probe, lens and sensor hardware, including additive manufacturing use cases. This functions across industrial construction of aerospace vehicle parts, testing and quality assurance. One aspect of the paperless supply


chain at Lockheed Martin UK is wire arc additive manufacturing (WAAM), utilised to digitally 3D print and repair parts made of various types of steel and other metals. Using electric arc robots as opposed to less sustainable powders with low build rates and smaller build volumes, wires are melted and molten metal is deposited in line with a predetermined pattern of layers. A rotating table and arm operating at the facility is estimated to complete production of parts in approximately six hours at a time. To determine whether part designs can be recycled in order to reduce waste, WAAM design capabilities can help ensure that said designs are fit for further purpose. Research projects using WAAM have been shown to reduce discarded waste materials by 55%. Build bay and stage production


progress documentation is maintained by production engineers on-site over computers and tablets, at macro level as well as site level. Queueing of the assemble-to-order


process is displayed in real time on screens, in the form of 3D model simulation that is generated by sensors. This paperless system is estimated to have cut 10 years of production time down to mere minutes. Across the aerospace sector, constant


updates and feedback from manufacturers to vendors early on in applications are vital to long-term success. According to Lopez, specialised function


teams for producing new algorithms also proves important: “Whether it’s classical, rules-based image processing, or it’s on the AI and deep learning side, we have people who learn about problems that our


“Companies are looking to develop their own applications and achieve specific use cases, including 3D, without requiring an army of programmers with really high machine


vision expertise” Sam Lopez, senior director at Zebra Technologies Canada


customers are seeing, and find innovative ways to solve those issues, which end users will then test and validate.” Machine vision and industrial scanning


software such as Zebra’s Aurora Focus operates on fixed industrial scanners and smart cameras, which come designed for specific tasks, including optical character recognition (OCR), and presence and absence vision inspection.


Metrology compliance in inspection In aerospace, all components of a vehicle need to be accurately measured first time, to meet industry guidelines as well as ensuring long-term fitness for purpose. As well as metals, composite materials are being increasingly utilised for long-term durability and malleability. However, composite resources can


prove more expensive in the long run than metals. To ensure optimal conditioning of materials, sensor data is converted into AI- generated images that demonstrate where more work is required, ensuring accurate measurements and less waste. FARO articulated arms, containing a point probe and laser scanners measuring down to 50 microns, are utilised by Lockheed Martin UK to carry out dimensional verification against industry guidelines, to ensure compliance. The coordinate measuring machine (CMM) determines and records specific locations of the probe at each point via embedded encoders on each axis, in 3D space. From here, results are reported on OCR


software on a computer screen. This generates a vectorised image on screen, which is overlaid over a 3D model that details any areas that need amendment. Areas that are good to go show up as green. While six-axis (for probing applications)


or seven-axis (for additional articulation) configuration arms are common on the market, FARO offers an eight-axis


> AUGUST/SEPTEMBER 2024 IMAGING AND MACHINE VISION EUROPE 29


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39