1,721,105 research outputs found
Flexible Task Execution and Attentional Regulations in Human-Robot Interaction
No full textA robotic system that interacts with humans is expected to flexibly execute structured cooperative tasks while reacting to unexpected events and behaviors. In this paper, we face these issues presenting a framework that integrates cognitive control, executive attention, and hierarchical plan execution. In the proposed approach, the execution of structured tasks is guided by top-down (task-oriented) and bottom-up (stimuli-driven) attentional processes that affect behavior selection and activation, while resolving conflicts and decisional impasses. Specifically, attention is here deployed to stimulate the activations of multiple hierarchical behaviors orienting them toward the execution of finalized and interactive activities. On the other hand, this framework allows a human to indirectly and smoothly influence the robotic task execution exploiting attention manipulation. We provide an overview of the overall system architecture discussing the framework at work in different case studies. In particular, we show that multiple concurrent tasks can be effectively orchestrated and interleaved in a flexible manner; moreover, in a human-robot interaction setting, we test and assess the effectiveness of attention manipulation for interactive plan guidance
Combined Text-Visual Attention Models for Robot Task Learning and Execution
No full textIn this work, we explore the interplay between text and visual attention mechanisms in a robot reinforcement learning setting, where robotic tasks are conveyed through natural language instructions. Specifically, we propose a novel approach aimed at enhancing robot task learning and execution by leveraging an integrated multimodal attention model that associates task-relevant environmental features with related words in the natural language mission text. We illustrate the overall framework architecture along with the learning process, emphasizing the interaction between textual and visual feature-based attention mechanisms. The method is trained in MiniGrid environments using the Proximal Policy Optimization algorithm, and its performance is evaluated by comparing the proposed architecture with a baseline that lacks attentional mechanisms. Experimental results demonstrate the efficacy of the approach also highlighting its potential in behavior transparency
Incremental Learning of Robotic Manipulation Tasks through Virtual Reality Demonstrations
No full textWe propose an incremental, modular, and extensible method for learning robotic manipulation tasks using a limited number of demonstrations provided in Virtual Reality, while assuming minimal prior information about the objects to be manipulated. The developed framework enables an incremental training process in which the operator first demonstrates specialized tasks to the robotic system, and subsequently more complex tasks, exploiting the skills learned during the previous phases. We illustrate and discuss the method at work considering picking tasks performed by manipulators equipped with multi-fingered sensorized hands. The experimental evaluation highlights the feasibility and advantage of the proposed method, particularly in terms of modularity, low number of demonstrations, and reliability of the trained system
Plan Execution and Attentional Regulations for Flexible Human-Robot Interaction
No full textFlexible execution of structured tasks is a relevant issue in cognitive robotics and human-robot interaction. In this work, we address this problem presenting a framework that integrates planning and attentional regulations for flexible and interactive execution of human-robot cooperative tasks. In the proposed approach, attentional top-down and bottom-up mechanisms are deployed to guide the execution of generated hierarchical plans while managing conflicts and decisional impasses. We provide an overview of the proposed framework discussing the system at work in different scenarios. In particular, we focus on flexible execution of multiple plans and interactive plan execution guided by attentional manipulation
Game−Theoretic Agent Programming in Golog
Full text linkWe present the agent programming language GTGolog, which integrates explicit agent programming in Golog with game-theoretic multi-agent planning in Markov games. It is a generalization of DTGolog to a multi-agent setting, where we have two competing single agents or two competing teams of agents. The language allows for specifying a control program for a single agent or a team of agents in a high-level logical language. The control program is then completed by an interpreter in an optimal way against another single agent or another team of agents, by viewing it as a generalization of a Markov game, and computing a Nash strategy. We illustrate the usefulness of this approach along a robotic soccer example
Model-based Executive Control through Reactive Planning for Autonomous Rovers
No full textThis paper reports on the design and implementation of a real-time executive for a mobile rover that uses a model-based, declarative approach. The control system is based on the Intelligent Distributed Execution Architecture (IDEA), an approach to planning and execution that provides a unified representational and computational framework for an autonomous agent. The basic hypothesis of IDEA is that a large control system can be structured as a collection of interacting agents, each with the same fundamental structure. We show that planning and real-time response are compatible if the executive minimizes the size of the planning problem. We detail the implementation of this approach on an exploration rover (Gromit an RWI ATRV Junior at NASA Ames) presenting different IDEA controllers of the same domain and comparing them with more classical approaches. We demonstrate that the approach is scalable to complex coordination of functional modules needed for autonomous navigation and exploration
- …
