Self-Modulated Adaptive Robotic Deposition: an Application to the Aerospace Industry

In modern industrial applications, the ability of robots to learn and adapt their motion is increasingly necessary. This is particularly true in tasks such as welding and material deposition, where the robot must plan and execute a path for a wide range of target scenarios. The current method of setting up these applications requires expert operators to manually generate a robot path for each part, which is a time-consuming process that can take at least one day to complete for each new part. In the case of deposition tasks, several activities must be carried out, including generating a deposition path in the CAD environment, converting the path into robot instructions, defining collisions-free robot motion, testing for quality assessment, and updating the path for desired deposition quality. To reduce setup time and standardize the task, there is a high demand for automating the path generation to achieve the desired quality. This paper focuses on automatic sealant deposition for the assembly of aerospace parts, which requires adapting the deposition path based on the target part geometry and material properties. The proposed Self-Modulated Adaptive Robotic Deposition (SMART DEPOSITION) tool offers a solution for generating and adapting robot motion in response to the deposition requirements of a target part. To accomplish this, the approach first analyzes the input part geometry to generate a grid of points. Then, an optimal deposition path is computed by exploiting the traveling salesman problem, using multiple approaches such as the nearest neighbor, a two-opt algorithm, and simulated annealing. The generated path can then be used to plan and execute the robot motion, resulting in the desired deposition quality with a standardized approach. The simulation results demonstrate the effectiveness of the proposed approach in optimizing the deposition path for different parts with varying geometries.