RoboPlanner : a pragmatic task planning framework for autonomous robots

Robotic automation has proliferated various industrial deployments including manufacturing, retail warehousing and logistics supply chains. In order for robots to advance to the next stage of cognitive autonomy, a robust framework for planning, execution and adaptation is needed. While there have been advances in abstract automated planning systems, they are still ill-suited to be applied within runtime robotic executions, which take place in uncertain environments. In this study, the authors provide a deliberative robotic planning and simulated execution framework called RoboPlanner that provides a pragmatic integration of automated planning, orchestration and adaptive deployments. This is coupled with an execution monitor and plan repair module, that allows reconfiguration to various template actions with runtime changes. Structured rules for re-planning in the case of state changes, unforeseen obstacles or execution failures are provided. They demonstrate their simulation framework on a realistic example of mobile pick & delivery robots in Industry 4.0 warehouses, that plan, execute, adapt and re-plan in sync with a knowledge base.