Planning and control model for robotic bridge painting

The objective of this research has been to develop and test a planning and control model for robotic bridge painting. The key element in this model is the Intelligent Painting Process Planner (IP$\sp3).$ Using the parameters provided by the IP$\sp3,$ the robotic bridge painting system will perform the painting operations with specified quality at lowest cost. In order to develop the IP$\sp3,$ the first step was to create bridge features which can be used to represent steel bridges. Thus, painting operations on the bridge can be decomposed into painting on each bridge feature. Second, laboratory experiments were conducted to investigate the aspects critical to the development of the IP$\sp3.$ Three questions under investigation were (1) what are the parameters which affect the coating quality? (2) what parameter values should be used to achieve the required quality? and (3) what are the coating thickness distribution functions for different parameters? The knowledge gained from the lab experiments is the fundament for building the IP$\sp3.$ The IP$\sp3$ is configured in two major components. They are (1) the process plan generator, and (2) the parameters editor. Eight process plan parameters are provided by the generator. These are: (1) the air pressure, (2) the fluid pressure, (3) the spray gun pitch angle, (4) the spray gun roll angle, (5) the distance, (6) the moving speed, (7) the number of runs, and (8) the reposition distance between two adjacent spray runs. The parameter values are selected based on two criteria: (1) satisfying the required quality, and (2) minimizing the total cost of the spray painting. Field experiments were conducted to test the developed model. Results of the experiments have demonstrated that not only is the robotic bridge painting feasible, but the required quality is also achievable.