[3B2] Flexible robotic programming for complex high-value inspection

M Vasilev¹, R Vithanage¹, K Burnham², C Loukas¹, H Gover¹, E Mohseni¹, R Zimermann¹, D Lines¹, Y Javadi¹, C MacLeod¹, G Pierce¹, A Gachagan¹, S Williams³ and J Ding³
¹University of Strathclyde, UK
²National Manufacturing Institute Scotland (NMIS), UK
³Cranfield University, UK 

Ultrasonic inspection is a regulatory requirement for the inspection of high-value components and is often viewed as the bottleneck of the manufacturing process. There is, therefore, the industrial drive towards highly customised automation technologies to perform such inspections. Despite the recent progress made by automated robotic systems, they still lack the fundamental technology to perform fully autonomous non-destructive testing and evaluation (NDT&E) without human intervention. Therefore, this research presents and discusses the feasibility and advantages of using flexible robotic approaches in NDT&E to exploit human skills, such as dexterity, know-how, analysis and advanced decision-making abilities, and robotic capabilities, such as strength, accuracy and repeatability, to inspect high-integrity components. The body of work presented herein uses KUKA industrial robots together with phased array ultrasound testing (PAUT) roller probe technology to inspect carbon fibre-reinforced polymer (CFRP) composite structures and additively manufactured metal components.

The flexible robotic programming technique discussed in this research aims to provide in-situ robotic programming capabilities, enabling human inspectors to program these systems by task demonstration. This is advantageous when there is little or no prior knowledge of advanced robotics or part geometries at hand. Furthermore, the proposed robotic inspection approaches provide real-time trajectory optimisation through constant contact force for improved inspection results and to accommodate as-built part tolerances. In contrast to conventional automated inspection cells, the proposed flexible inspection technology will enable easily reconfigurable work cells to inspect a variety of high-value components such as aerospace and defence structures, wind turbine blades and complex pipe geometries and pressure vessels.

These systems will also increase accuracy and throughput by combining unique human skills and robotic capabilities. Furthermore, it is expected that these systems will improve safety and ergonomics by utilising automation to perform mundane bulk inspections of large surfaces.