Suspending the task Sample Clauses

Suspending the task. In some situation, the robot needs to suspend the task, because the user has stopped following it. In this case, the robot should estimate if this suspension of the collaborative scenario is temporary or permanent, and in the latter case abandon the task. We estimate this information using the Suspend Model and the activity areas from Situation Assessment. We link activity areas to the maximum time we expect that the user will be involved in the activity, and with a set of proactive actions that the robot can choose to execute. In this paper, we investigated a single possible proactive behavior: giving information. In this case, if we detect that the user has stopped following because he is looking at a touristic sight, or at ICT-FP7-600877-▇▇▇▇▇▇▇ Deliverable D5.6 an information screen, the robot can try to engage him and offer related information. At the moment, we just propose a simple routine-based framework for this behavior, and plan to further study it in the future. We believe that the solution of this problem could be rich, and that the robot should estimate the reaction of the user during the execution of its proactive behavior, in order to be able to interrupt if he doesn’t want to be helped or to resume the original task if he is satisfied by the robot’s actions. We don’t want the robot to be stuck for a long time waiting for a person. If there is a small amount of time to reach the destination, or the user is engaged in the activity for a longer period of time than the one predicted, the Suspend Model can issue a warning action, and eventually abandon the task if the person doesn’t start following it again. Sometimes users will stop following without an apparent reason, perhaps outside any activity area. In this case the robot will still allow them some time before issuing a warning and eventually abandoning the task. [1] D5.5: Task planner and robot supervision system early prototype. [2] ▇▇▇▇▇▇▇ second year report. [3] ▇▇▇▇▇-▇▇´ ▇▇▇, M., ▇▇▇▇▇▇, ▇., ▇▇▇▇▇▇, O., AND ▇▇▇▇▇▇▇▇▇▇, F. A closer look at momdps. In Tools with Artificial Intelligence (ICTAI), 2010 22nd IEEE International Confer- ence on (2010), vol. 2, IEEE, pp. 197–204. [4] ▇▇▇▇▇, M., ▇▇▇▇▇▇▇▇▇▇, H., ▇▇▇▇▇▇▇, ▇., AND ▇▇▇▇▇, ▇. ▇▇ adaptive and proactive human-aware robot guide. In Social Robotics. Springer, 2015, pp. 194–203. [5] ▇▇▇▇▇▇, ▇., ▇▇▇, ▇., AND ▇▇▇▇▇, S. A hierarchical approach to pomdp planning and execu- tion. In Workshop on hierarchy and memory in reinforcement learning ...