Esther Luna Colombini

This thesis presents a computational framework for intrinsically motivated autonomous agents, formalizing intrinsic motivation (IM) as a multi-objective reinforcement learning problem integrating drive regulation, hedonic valuation, hierarchical need prioritization, and Theory of Mind, enabling adaptive, long-term decision-making and socially aware interaction. Validated on simulated and physical robotic platforms, the framework demonstrates improved behavioral stability, policy reshaping via hedonic modulation, and enhanced cooperation among agents with compatible motivational profiles and one altruistic agent. This work extends beyond robotics, offering a computational model of IM that bridges cognitive science and autonomous decision-making.

A aquisição de habilidades motoras e cognitivas em robôs humanoides ainda depende fortemente de engenharia de tarefas e instruções explícitas, restringindo-os a cenários limitados e pré-definidos. Nosso trabalho, o framework DreamerRL, supera essas limitações ao permitir que o agente aprenda habilidades de forma totalmente autônoma, guiado unicamente por sua curiosidade em explorar e entender o ambiente. Em experimentos realizados com o robô humanoide NAO, demonstramos que o DreamerRL avança o estado da arte ao possibilitar a emergência espontânea de comportamentos manipulativos complexos e habilidades cognitivas essenciais, tipicamente observadas em crianças de até três anos, sem a necessidade de intensa engenharia humana.

This research explores how robots can autonomously adapt to complex environments using a human-inspired cognitive development framework. By integrating intrinsic motivation and decision-making, robots employed reinforcement learning to balance internal states. An extended model introduced the concept of pleasure, enabling context-aware prioritization of needs and preferences. A child-inspired cognitive architecture guided learning, interaction, and survival, giving rise to distinct robot personalities that shaped human engagement. Building on this, an advanced model with Theory of Mind allowed two robots to collaborate through mutual understanding of needs. We validated our models on both simulated and real robots with varied architectures.

This study explores how humans and cognitive robots with different value systems and motivations understand each other's needs in free-form interactions. We developed a cognitive architecture that links sensing and perception to internal motivation and an intrinsic value system for determining actions. Inspired by young children's needs, this architecture includes three drives: learning, interaction, and recharging, each with varying dependence on the human partner. We aimed to assess how experimentally changing the importance of these drives within a fixed architecture affects interaction dynamics with human partners (acting as caregivers) and their understanding of the robot's needs. By adjusting the learning and interaction drives, we created two robot profiles: Playful, which prioritizes environmental exploration and playfulness to reduce boredom, and Social, which focuses on social interaction through touch and visual contact to increase comfort. Our findings show that changing the importance of these drives produces distinct behaviors and human perceptions. Robot behaviors matched their profiles, and participants adapted their responses accordingly. Participants identified and attributed distinct traits to each robot without knowing the specific profiles. Despite variability among human partners, the robots, especially the playful one, were generally well understood by most participants.

The ability to communicate with robots using natural language is a significant step forward in human-robot interaction. However, accurately translating verbal commands into physical actions is promising, but still presents challenges. Current approaches require large datasets to train the models and are limited to robots with a maximum of 6 degrees of freedom. To address these issues, we propose a framework called InstructRobot that maps natural language instructions into robot motion without requiring the construction of large datasets or prior knowledge of the robot's kinematics model. InstructRobot employs a reinforcement learning algorithm that enables joint learning of language representations and inverse kinematics model, simplifying the entire learning process. The proposed framework is validated using a complex robot with 26 revolute joints in object manipulation tasks, demonstrating its robustness and adaptability in realistic environments. The framework can be applied to any task or domain where datasets are scarce and difficult to create, making it an intuitive and accessible solution to the challenges of training robots using linguistic communication. Open source code for the InstructRobot framework and experiments can be accessed at https://github.com/icleveston/InstructRobot.

A Mostra Nacional de Robótica (MNR) é uma das mostras científicas de caráter nacional apoiadas pelo governo federal no Brasil. Sua temática é centrada na automação e na robótica, representantes inatos das novas tecnologias, elementos com forte capacidade para estimular o imaginário das pessoas e elementos fortemente integrados com processos modernos de ensino-aprendizagem, dialogando diretamente com diversas disciplinas dos ciclos fundamental, médio e técnico. O formato adotado pela MNR busca reunir elementos de divulgação científica típicos das atividades museológica, literária e de exposição, aliada ao uso da multimídia como estratégia de aproximação dos jovens, em sintonia com as demandas da pós-modernidade. O presente artigo apresenta um retrospecto da experiência de pouco mais de uma década de organização da MNR (2011-2022), discute os principais números e resultados alcançados, e faz uma análise da eficácia da ação na divulgação científica e como política pública, e aponta caminhos para o crescimento desta atividade no Brasil.

Data derived from the realm of the social sciences is often produced in digital text form, which motivates its use as a source for natural language processing methods. Researchers and practitioners have developed and relied on artificial intelligence techniques to collect, process, and analyze documents in the legal field, especially for tasks such as text summarization and classification. While increasing procedural efficiency is often the primary motivation behind natural language processing in the field, several works have proposed solutions for human rights-related issues, such as assessment of public policy and institutional social settings. One such issue is the presence of gender biases in court decisions, which has been largely studied in social sciences fields; biased institutional responses to gender-based violence are a violation of international human rights dispositions since they prevent gender minorities from accessing rights and hamper their dignity. Natural language processing-based approaches can help detect these biases on a larger scale. Still, the development and use of such tools require researchers and practitioners to be mindful of legal and ethical aspects concerning data sharing and use, reproducibility, domain expertise, and value-charged choices. In this work, we (a) present an experimental framework developed to automatically detect gender biases in court decisions issued in Brazilian Portuguese and (b) describe and elaborate on features we identify to be critical in such a technology, given its proposed use as a support tool for research and assessment of court activity.