High-Resolution Digital Tactile Sensors for Streamlined Maintenance Workflows

An essential technique in aircraft maintenance is the use of visual inspection to detect flaws and corrosion in critical aircraft sections and components. However, the lack of digital data from most of these methods limits the ability to analyze and learn from maintenance data over time. There is a growing need to modernize aircraft maintenance processes through digitization of inspection workflows and archiving of accurate digital measurements of corrosion, cracks, and other types of wear. Accurate three-dimensional measurements of components of any material used in an aircraft is a fundamental technology that unlocks the potential of a fully digitized inspection process.GelSight’s handheld 3D measurement systems are widely used within the commercial aerospace industry for measuring the surface geometry of scratches, damage and other indications on components with precision down to 0.0001”. The work in this proposal will extend the GelSight measurement platform with two capabilities to streamline its use within maintenance workflows. The first extension is to track the location of GelSight measurements relative to the component or aircraft. This location information is essential to provide context to the surface damage and wear measurements captured by GelSight systems. The second extension to the GelSight platform is to develop a measurement probe with connectivity and software APIs to facilitate integration into robotics and automation platforms. This improvement will enable automated inspections to build rich datasets for training machine learning models. The successful completion of both goals in this proposal will enable the capture of digital information that will provide deeper insights into wear patterns, how damage changes over time, and locations on components or aircraft that are prone to different types of wear. This information is essential for modernizing aircraft maintenance workflows towards accurate on-condition and predictive maintenance decisions.