Generating Hand-Written Symbols With Trajectory Planning Using A Robotic Arm

In this paper an evolutionary-based pattern generator to draw lines, curves, and shapes using a robotic arm without explicit instructions, is presented. Drawing is considered to be denoting a symbol with continuous curves and lines, which is different from the work of a printer. First, we consider drawing the digits 0 to 9 in the simulation environment with a robotic arm. Next, the ability of the method to learn drawing more complicated patterns is studied. The method is applied to control the irb4600 robotic arm for drawing patterns, in the Webots simulation environment. Our proposed method is compared with some other relevant methods the advantages and disadvantages of each approach were examined. The advantage of our algorithm is that it allows us to draw the desired shapes by the robot without prior training and the need for large amounts of data. The results of this paper can greatly benefit the applications in the industry of autonomous cutting, welding, drawing, and industrial design. We were able to successfully draw various types of shapes and symbols in the simulation and generate an accurate (more than 94% across distinct runs) trajectory for our robot.