Sancho Oliveira | ISCTE - University Institute of Lisbon (ISCTE-IUL) (original) (raw)
Uploads
Papers by Sancho Oliveira
ACM SIGDOC European Chapter/ Eurosigdoc Workshop on Open Source and Design of Communication - OSDOC '10, 2010
Traditionally the teaching of programming is done by the usage of industrial tools (full blown ID... more Traditionally the teaching of programming is done by the usage of industrial tools (full blown IDE's), this paper will present the current development status of Botbeans, a visual programming language and tool that will enable its users to see the results of their programming in a tangible interface, where a robot (Lego Mindstorms NXT) will show the results of its programming while interacting with the real world.
Proceedings of the International Conference on Agents and Artificial Intelligence, 2015
Artificial Life 14: Proceedings of the Fourteenth International Conference on the Synthesis and Simulation of Living Systems, 2014
Artificial Life 14: Proceedings of the Fourteenth International Conference on the Synthesis and Simulation of Living Systems, 2014
Proceedings of the Workshop - Open Source and Design of Communication, OSDOC 2010, 2010
Abstract In this paper we propose an open source architecture, which enables the scientific commu... more Abstract In this paper we propose an open source architecture, which enables the scientific community to implement and execute highly demanding CPU tasks in distibuted way. The system main idea is to use idle time of a set of volunteers, and distribute in a smart and transparent way the work to the volunteers that perform the real work. Contrary to the existing solutions, all the deployment of the tasks, code and data, is done automatically by the system, allowing a very easy installation and submission process. This system is ...
We propose and evaluate a novel approach called Online Distributed NeuroEvolution of Augmenting T... more We propose and evaluate a novel approach called Online Distributed NeuroEvolution of Augmenting Topologies (odNEAT). odNEAT is a completely distributed evolutionary algorithm for online learning in groups of embodied agents such as robots. While previous approaches to online distributed evolution of neural controllers have been limited to the optimisation of weights, odNEAT evolves both weights and network topology. We demonstrate odNEAT through a series of simulation-based experiments in which a group of e-puck-like robots must perform an aggregation task. Our results show that robots are capable of evolving effective aggregation strategies and that sustainable behaviours evolve quickly. We show that odNEAT approximates the performance of rtNEAT, a similar but centralised method. We also analyse the contribution of each algorithmic component on the performance through a series of ablation studies.
Lecture Notes in Computer Science, 2012
We present a novel methodology for the synthesis of behavioral control for real robotic hardware.... more We present a novel methodology for the synthesis of behavioral control for real robotic hardware. In our approach, neural controllers decide when different preprogrammed behaviors should be active during task execution. We evaluate our approach in a double T-maze task carried out by an e-puck robot. We compare results obtained in our setup with results obtained in a traditional evolutionary robotics setup where the neural controller has direct control over the robot's actuators. The results show that the combination of ...
Lecture Notes in Computer Science, 2011
Studies on social insects have demonstrated that complex, adaptive and self-organized behavior ca... more Studies on social insects have demonstrated that complex, adaptive and self-organized behavior can arise at the macroscopic level from relatively simple rules at the microscopic level. Several past studies in robotics and artificial life have focused on the evolution and understanding of the rules that give rise to a specific macroscopic behavior such as task allocation, communication or synchronization. In this study, we demonstrate how colonies of embodied agents can be evolved to display multiple complex macroscopic behaviors at ...
Journal of Intelligent & Robotic Systems, 2014
We propose an approach to the synthesis of hierarchical control systems comprising both evolved a... more We propose an approach to the synthesis of hierarchical control systems comprising both evolved and manually programmed control for autonomous robots. We recursively divide the goal task into subtasks until a solution can be evolved or until a solution can easily be programmed by hand. Hierarchical composition of behavior allows us to overcome the fundamental challenges that typically prevent evolutionary robotics from being applied to complex tasks: bootstrapping the evolutionary process, avoiding deception, and successfully transferring control evolved in simulation to real robotic hardware. We demonstrate the proposed approach by synthesizing control systems for two tasks whose complexity is beyond state of the art in evolutionary robotics. The first task is a rescue task in which all behaviors are evolved. The second task is a cleaning task in which evolved This work was partly supported by FCT -Fundação para a Ciência e Tecnologia (FCT) Portugal under grants SFRH/ BD/76438/2011, PEst-OE/EEI/LA0008/2013 and EXPL/ behaviors are combined with a manually programmed behavior that enables the robot to open doors in the environment. We demonstrate incremental transfer of evolved control from simulation to real robotic hardware, and we show how our approach allows for the reuse of behaviors in different tasks.
2012 IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL), 2012
In this paper, we demonstrate how an artificial neural network (ANN) based controller can be synt... more In this paper, we demonstrate how an artificial neural network (ANN) based controller can be synthesized for a complex task through hierarchical evolution and composition of behaviors. We demonstrate the approach in a task in which an e-puck robot has to find and rescue a teammate. The robot starts in a room with obstacles and the teammate is located in a double T-maze connected to the room. We divide the rescue task into different sub-tasks: (i) exit the room and enter the double T-maze, (ii) solve the maze to find the teammate, and (iii) guide the teammate safely to the initial room. We evolve controllers for each sub-task, and we combine the resulting controllers in a bottom-up fashion through additional evolutionary runs. We conduct evolution offline, in simulation, and we evaluate the highest performing controller on real robotic hardware. The controller achieves a task completion rate of more than 90% both in simulation and on real robotic hardware.
Proceedings of the International Conference on Agents and Artificial Intelligence, 2015
Proceedings of the Workshop on Open Source and Design of Communication - OSDOC '13, 2013
ACM SIGDOC European Chapter/ Eurosigdoc Workshop on Open Source and Design of Communication - OSDOC '10, 2010
Traditionally the teaching of programming is done by the usage of industrial tools (full blown ID... more Traditionally the teaching of programming is done by the usage of industrial tools (full blown IDE's), this paper will present the current development status of Botbeans, a visual programming language and tool that will enable its users to see the results of their programming in a tangible interface, where a robot (Lego Mindstorms NXT) will show the results of its programming while interacting with the real world.
Proceedings of the International Conference on Agents and Artificial Intelligence, 2015
Artificial Life 14: Proceedings of the Fourteenth International Conference on the Synthesis and Simulation of Living Systems, 2014
Artificial Life 14: Proceedings of the Fourteenth International Conference on the Synthesis and Simulation of Living Systems, 2014
Proceedings of the Workshop - Open Source and Design of Communication, OSDOC 2010, 2010
Abstract In this paper we propose an open source architecture, which enables the scientific commu... more Abstract In this paper we propose an open source architecture, which enables the scientific community to implement and execute highly demanding CPU tasks in distibuted way. The system main idea is to use idle time of a set of volunteers, and distribute in a smart and transparent way the work to the volunteers that perform the real work. Contrary to the existing solutions, all the deployment of the tasks, code and data, is done automatically by the system, allowing a very easy installation and submission process. This system is ...
We propose and evaluate a novel approach called Online Distributed NeuroEvolution of Augmenting T... more We propose and evaluate a novel approach called Online Distributed NeuroEvolution of Augmenting Topologies (odNEAT). odNEAT is a completely distributed evolutionary algorithm for online learning in groups of embodied agents such as robots. While previous approaches to online distributed evolution of neural controllers have been limited to the optimisation of weights, odNEAT evolves both weights and network topology. We demonstrate odNEAT through a series of simulation-based experiments in which a group of e-puck-like robots must perform an aggregation task. Our results show that robots are capable of evolving effective aggregation strategies and that sustainable behaviours evolve quickly. We show that odNEAT approximates the performance of rtNEAT, a similar but centralised method. We also analyse the contribution of each algorithmic component on the performance through a series of ablation studies.
Lecture Notes in Computer Science, 2012
We present a novel methodology for the synthesis of behavioral control for real robotic hardware.... more We present a novel methodology for the synthesis of behavioral control for real robotic hardware. In our approach, neural controllers decide when different preprogrammed behaviors should be active during task execution. We evaluate our approach in a double T-maze task carried out by an e-puck robot. We compare results obtained in our setup with results obtained in a traditional evolutionary robotics setup where the neural controller has direct control over the robot's actuators. The results show that the combination of ...
Lecture Notes in Computer Science, 2011
Studies on social insects have demonstrated that complex, adaptive and self-organized behavior ca... more Studies on social insects have demonstrated that complex, adaptive and self-organized behavior can arise at the macroscopic level from relatively simple rules at the microscopic level. Several past studies in robotics and artificial life have focused on the evolution and understanding of the rules that give rise to a specific macroscopic behavior such as task allocation, communication or synchronization. In this study, we demonstrate how colonies of embodied agents can be evolved to display multiple complex macroscopic behaviors at ...
Journal of Intelligent & Robotic Systems, 2014
We propose an approach to the synthesis of hierarchical control systems comprising both evolved a... more We propose an approach to the synthesis of hierarchical control systems comprising both evolved and manually programmed control for autonomous robots. We recursively divide the goal task into subtasks until a solution can be evolved or until a solution can easily be programmed by hand. Hierarchical composition of behavior allows us to overcome the fundamental challenges that typically prevent evolutionary robotics from being applied to complex tasks: bootstrapping the evolutionary process, avoiding deception, and successfully transferring control evolved in simulation to real robotic hardware. We demonstrate the proposed approach by synthesizing control systems for two tasks whose complexity is beyond state of the art in evolutionary robotics. The first task is a rescue task in which all behaviors are evolved. The second task is a cleaning task in which evolved This work was partly supported by FCT -Fundação para a Ciência e Tecnologia (FCT) Portugal under grants SFRH/ BD/76438/2011, PEst-OE/EEI/LA0008/2013 and EXPL/ behaviors are combined with a manually programmed behavior that enables the robot to open doors in the environment. We demonstrate incremental transfer of evolved control from simulation to real robotic hardware, and we show how our approach allows for the reuse of behaviors in different tasks.
2012 IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL), 2012
In this paper, we demonstrate how an artificial neural network (ANN) based controller can be synt... more In this paper, we demonstrate how an artificial neural network (ANN) based controller can be synthesized for a complex task through hierarchical evolution and composition of behaviors. We demonstrate the approach in a task in which an e-puck robot has to find and rescue a teammate. The robot starts in a room with obstacles and the teammate is located in a double T-maze connected to the room. We divide the rescue task into different sub-tasks: (i) exit the room and enter the double T-maze, (ii) solve the maze to find the teammate, and (iii) guide the teammate safely to the initial room. We evolve controllers for each sub-task, and we combine the resulting controllers in a bottom-up fashion through additional evolutionary runs. We conduct evolution offline, in simulation, and we evaluate the highest performing controller on real robotic hardware. The controller achieves a task completion rate of more than 90% both in simulation and on real robotic hardware.
Proceedings of the International Conference on Agents and Artificial Intelligence, 2015
Proceedings of the Workshop on Open Source and Design of Communication - OSDOC '13, 2013