AR Metrology

Augmented Reality (AR), although not a new concept, has only in recent years started to gain traction as an applicable, useful technology. The term 'augmented reality' dates back only to 1990, though the concept can actually be attributed to the author of the Wizard of Oz, who came up with electronic spectacles that display information over real-life visuals back in 1901. Back at the start of the 20^th century, it was pure imagination, but we are now in an age where AR is not only possible, it is accessible, applicable and tremendously useful. 

FARO is one company that has released an AR tool to assist in manufacturing quality processes. FARO's Visual Inspect AR tool is designed to work on handheld tablets, meaning it's intuitive, easy to use and extremely versatile due to its portability. One major benefit is of course that tablets with high picture quality cameras are readily available on the market for relatively low costs. This means that no specialised, expensive tools are required to run the AR software. Visual Inspect can assist in multiple ways, by overlaying data onto the camera feed. CAD-to-part comparison is one of the main applications, but others are mentioned as well, such as virtual part validation, component alignment & assembly and more. 

A recent paper for the Engineering and Technology International Journal of Aerospace and Mechanical Engineering of the World Academy of Science, Nuran Bahar and Mehmet A. Akcayol highlighted the value of using AR specifically in the aerospace sector. The full paper is called Improvement of Realization Quality of Aerospace Products Using Augmented Reality Technology and you can read it here, though I will summarise below. 

The paper opens up with mentioning that AR has come into use in numerous areas and that the aviation industry is one where it can be very beneficial: "If we look from the viewpoint of aviation industry, technical employees have to be very experienced at areas of system design, implementation and assembly inspection." Having to consult external resources such as digital mock ups (DMU), detracts from these employees' concentration and severely slows down production. Aircraft products are only becoming more complex as time goes on, which means this issue has the potential to escalate. 

 Augmented reality, and also virtual reality, have become viable options to streamline Computer Aided Manufacturing (CAM), or processes such as referencing 3D CAD data. An AR based tool-chain was presented for the paper, "which allows matching the CAD data with real mock-ups. In [the] solution, the CAD model and the physical mock-up are aligned in a live video sequence." The user can then align the virtual and real-life geometry visually in order to check if everything is correct. 

The example given is the inspection of the upper shell of an aircraft. This is, of course, one of multiple stages at which checks are done to ensure that assembly has been completed correctly. Whereas formerly, the checks would have to be a fully manual-visual inspection, now AR assists in this task. The technical controller team checks whether there are any missing parts, whether brackets are in the right position/direction, and should a bracket have electrical bounding, this needs to be checked as well. If all these components have the correct configuration according to the technical design documents, the examination process is concluded successfully. Fig. 1 shows this these different components. 

Marker based tracking & Computer Vision (CV) based tracking

In marker based tracking, fiducial markers (essentially a point of reference for a camera) is used to determine and detect the position of the camera in relation to the marker. This brings with it some disadvantages: attachment of the markers to the object that is to be measured can be a challenge. For a shape that is as diverse and complicated as the aircraft panel shown above, this would mean having an abundance of markers placed on the object. This would prove to be very time-consuming, which in turn would reduce the attractiveness of using AR as opposed to traditional methods. After all, time-saving is one of the main attractive properties of AR for industrial applications. 

Marker-less tracking methods are therefore an attractive option to use. One of these is Simultaneous Localisation and Mapping (SLAM). SLAM is the the "computational problem of constructing or updating a map of an unknown environment while simultaneously keeping track of an agent's location within it. [...] SLAM consists of multiple parts; Landmark extraction, data association, state estimation, state update and landmark update. [...] The goal of the process is to use the environment to update the position of the robot." Essentially, it's a move towards semi-autonomous inspection, carried out by a robot. When reading this, I instantly though of Boston Dynamics' Spot robot, which can be equipped with different scanners to autonomously explore locations and, I imagine, will be used in the future for applications just like this. 

In their conclusion, Nuran Bahar and Mehmet A. Akcayol mention that the "Developed application is designed to reduce the assembly inspection time of tens of thousands [of] brackets that may hold pipes, wire bundles, electrical bounding's [sic] in place in the aircraft part and find out wrongly positioned, damaged or missing brackets." It also minimises the errors that can occur when mounting brackets, improves workflow efficiency and reduce labour costs by reducing the amount of defects. In the future, they mention, heads-up-display (HUD) goggles could be used to streamline the task even further, keeping engineers' hands free while conducting inspections. 

AR is already in use in many ways, but it's certainly one of those technologies that will become ubiquitous within the next few years as a move to Industry 4.0.