- Download 7
- File Size 185 KB
- File Count 1
- Create Date October 5, 2021
- Last Updated October 5, 2021
D3.1 Report and software for human-aware task planning and scheduling in HRC systems (Module#6) - ABSTRACT
This deliverable aims at describing the scientific and technical results achieved within the task T3.1 “Dynamic planning and scheduling for robot tasks considering human needs” within WP3. This task has focused on designing and developing a human-aware task planning and scheduling system suitable for Human-Robot Collaboration. The main feature of this system is to deploy flexible and adaptable task control of a collaborative robot leveraging the knowledge about the production process, the robot, the operator and the working environment. Namely, all the information represented in the Knowledge Base and/or coming from the Environment Cognition (WP2, T2.1, T2.2.) are exploited to generate task plans tailored for the specific domain of work, the environmental context and the human features.
This module has two main contributors:
- CNR-ISTC, to design and develop a dynamic task planning and scheduling system responsible for continuously create an updated task plan. This system manages tasks assignment balancing human-robot coordination while increasing robot operational efficiency and preserving operator safety. The system is also integrated with the motion planning module to deploy a synergic task and motion planning approach to HRC control. Particular attention is given to maintenance of safety critical properties of the current task plan using a combination of formal techniques for Verification & Validation within the background of CNR-ISTC. To achieve such a goal, CNR-ISTC will develop a framework around the main modules of the task planning and scheduling system implementing a reactive control loop that guarantees real time control.
- The dynamic task planning and scheduling system is also enhanced by RWTH with additional features. The workspace cognition is translated in a module developed by RWTH into reliable and simple decision values to be used as triggers by the task planning and scheduling system. Therefore, in a further module, the decision values probabilistic decisions on the planning strategy to be followed will be derived. Further on, the planning approach is extended by RWTH towards parallel planning of main and alternative paths. Namely, in addition to the main plan, secondary (i.e., alternative) plans will be available such that in case that the main plan cannot be followed and there is no time for re-planning, special smoothing algorithms will ensure the change to one of the secondary plans.