International Conference on Automated Planning and Scheduling
ProSeqqo: Open-source Solver for Robotic Task Sequencing
ProSeqqo is a generic, open-source solver for process planning and sequencing in industrial robotics. It captures planning problems from diverse applications where task sequencing can be coupled with different types of discrete choices on how the tasks are performed: the selection of the inverse kinematic (IK) solution for a task originally defined in task space; choosing the direction of the motion in cutting or arc welding; or selecting the entry and exit points when following a closed contour. The planning problem can be declared in arbitrary dimensions, e.g., in the 3D task space or in the 6D robot joint configuration space. Moreover, the optimization criterion can be selected from a large set of predefined cost functions, such as minimizing the Euclidean travel distance, minimizing the travel time assuming limited robot joint velocities and accelerations, or a custom cost matrix can be defined. Additional side constraints, such as precedences or resource changeovers can also be captured. The planning problem can be defined using an easily readable problem definition language. This description is transformed into a generalized travelling salesman problem (GTSP) with precedence constraints, and then solved using the algorithms of the Google OR-tools VRP library. The demo video is available at https://youtu.be/vK53AcCoVT8 . Source codes and examples can be downloaded from https://github.com/sztaki-hu/ProSeqqo .