Adrián Romero-Garcés

Audio-visual perception system for a humanoid robotic head.

One of the main issues within the field of social robotics is to endow robots with the ability to direct attention to people with whom they are interacting. Different approaches follow bio-inspired mechanisms, merging audio and visual cues to localize a person using multiple sensors. However, most of these fusion mechanisms have been used in fixed systems, such as those used in video-conference rooms, and thus, they may incur difficulties when constrained to the sensors with which a robot can be equipped. Besides, within the scope of interactive autonomous robots, there is a lack in terms of evaluating the benefits of audio-visual attention mechanisms, compared to only audio or visual approaches, in real scenarios. Most of the tests conducted have been within controlled environments, at short distances and/or with off-line performance measurements. With the goal of demonstrating the benefit of fusing sensory information with a Bayes inference for interactive robotics, this paper presents a system for localizing a person by processing visual and audio data. Moreover, the performance of this system is evaluated and compared via considering the technical limitations of unimodal systems. The experiments show the promise of the proposed approach for the proactive detection and tracking of speakers in a human-robot interactive framework.

A DDS-based middleware for quality-of-service and high-performance networked robotics

Social robots must adapt to dynamic environments, human interaction partners and challenging new stringent tasks. Their inner software is usually distributed and should be designed and deployed carefully because slight changes in the robots requirements can have an important impact not only on the existing source code but also on the resulting performance at run‐time. This paper proposes an implementation of this inner software using a new lightweight middleware for networked robotics called Nerve. The principal novelty this middleware has with respect to other state‐of‐the‐art approaches is that it guarantees both scalability and QoS, which are key requirements for real‐time robotics software. The benefits of Nerve have been proved through its use in two key components of the cognitive system of a social robot: (i) the visual attention mechanism, used to extract relevant data from perceived images; and (ii) a robot learning by imitation control architecture that allows the social robot to be taught by people using natural demonstrations (i.e. using the same communication channels that would be used to teach people). Nerve makes use of existing patterns for networked applications together with the recent Data Distribution Service specification, which is a publish/subscribe standard for real‐time and distributed systems that provides a wide set of QoS policies. In this paper, these different QoS policies have been applied carefully to achieve the best performance of the target robot. Copyright

Testing a Fully Autonomous Robotic Salesman in Real Scenarios

Over the past decades, the number of robots deployed in museums, trade shows and exhibitions have grown steadily. This new application domain has become a key research topic in the robotics community. Therefore, new robots are designed to interact with people in these domains, using natural and intuitive channels. Visual perception and speech processing have to be considered for these robots, as they should be able to detect people in their environment, recognize their degree of accessibility and engage them in social conversations. They also need to safely navigate around dynamic, uncontrolled environments. They must be equipped with planning and learning components, that allow them to adapt to different scenarios. Finally, they must attract the attention of the people, be kind and safe to interact with. In this paper, we describe our experience with Gualzru, a salesman robot endowed with the cognitive architecture RoboCog. This architecture synchronizes all previous processes in a social robot, using a common inner representation as the core of the system. The robot has been tested in crowded, public daily life environments, where it interacted with people that had never seen it before nor had a clue about its functionality. Experimental results presented in this paper demonstrate the capabilities of the robot and its limitations in these real scenarios, and define future improvement actions.

Improving a robotics framework with real-time and high-performance features

Middleware has a key role in modern and object-oriented robotics frameworks, which aim at developing reusable, scalable and maintainable systems using different platforms and programming languages. However, complex robotics software falls into the category of distributed real-time systems with stringent requirements in terms of throughput, latency and jitter. This paper introduces and analyzes a methodology to improve an existing robotics framework with real-time and high-performance features using a recently adopted standard: the Data Distribution Service (DDS).

Recipes for designing high-performance and robust software for robots

Until now, high-performance has been the main objective in software for robotics and, as a result, the ad-hoc implementations have been optimized for specific hardware and platforms. Nevertheless, there is a renewed interest in designing robot control architectures to be reusable and maintainable as possible, so that existing software modules can be adapted to new platforms and requirements, thus reducing the cost and time-to-market of the complete system. This paper presents the most relevant conclusions of a joint-work between researchers from the telecommunications world (a domain with stringent requirements for distributed and real-time embedded systems) and researchers in the field of robotics. The challenge consisted of identifying the best practices and tools currently available in software engineering for embedded systems and protocols in order to define a precise methodology for the design of a high-performance and robust software control architecture of a robot. We outline the problems detected in current software developed for robots and then propose solutions to them.

A Model-Driven Approach to Enable Adaptive QoS in DDS-Based Middleware

Critical and distributed systems need to be reliable and comply with the required performance at run-time. In this vein, data distribution service for real-time systems (DDS) provides developers with highly configurable middleware to control the end-to-end quality of service (QoS) of the applications through a wide range of attributes. However, dynamic and unpredictable environment pose a major challenge to these systems as their workload and resources may fluctuate significantly in time depending on the execution context. Developers usually find it difficult to choose and apply the right DDS QoS attributes, as once selected, they remain fixed during the whole execution of the system. They do not automatically change according to the execution context, e.g., to meet nonfunctional requirements related to performance or resource consumption. Moreover, changing the QoS attributes at run-time may lead to incompatibilities, since the configuration used by the different participants needs to be mutually consistent. In this paper, we propose a model-driven approach that enables the safe, automatic, and transparent adaptation of the QoS attributes in DDS-based middleware, providing the best performance possible within the available resources at run-time. An example in robotics is presented to demonstrate the feasibility and the benefits of our proposal.

A Navigation Agent for Mobile Manipulators

Robot navigation and manipulation in partially known indoor environments is usually organized as two complementary activities, local displacement control and global path planning. Both activities have to be connected across different space and time scales in order to obtain a smooth and responsive system that follows the path and adapts to the unforeseen situations imposed by the real world. There is not a clear consensus in how to do this and some important problems are still open. In this paper we present the first steps towards a new navigation agent controlling both the robot’s base and the arm. We address several of theses problems in the design of this agent, including robust localization integrating several information sources, incremental learning of free navigation and manipulation space, hand visual servoing in camera space to reduce backslash and calibration errors, and internal path representation as an elastic band that is projected to the real world through measurements of the sensors. A set of experiments are presented with the robot Ursus in real and simulated scenarios showing some encouraging results.

Integrating the users in the design of a robot for making Comprehensive Geriatric Assessments (CGA) to elderly people in care centers

Comprehensive Geriatric Assessment (CGA) is a multidimensional and multidisciplinary diagnostic instrument that helps provide personalized care to the elderly, by evaluating their physical and mental state. In a social and economic context of growing ageing populations, medical experts can save time and effort if provided with interactive tools to efficiently assist them in doing CGAs, managing standardized tests or data collection. Recent research proposes the use of social robots as the central part of these tools. These robots must be able to unfold all functionalities that questionnaires or motion-based tests require, including natural language, face tracking and monitoring, human motion capture and so on. But another issue is the robots acceptability and trust by the end-users, both patients (elderly people) and clinicians: the robot needs to be able to engage with the patients during the interaction sessions, and must be perceived as a useful and efficient tool by the clinicians. This paper presents the acquisition of new user requirements for CLARC, through participatory and user-centered design approach, to inform the improvement of both interface and interaction. Thirty eight persons (elderly people, caregivers and health professionals) were involved in the design process of CLARC, based on user-centered methods and techniques of Human-Computer Interaction discipline.

CLARC: A Cognitive Robot for Helping Geriatric Doctors in Real Scenarios

Comprehensive Geriatric Assessment (CGA) is an integrated clinical process to evaluate the frailty of elderly persons in order to create therapy plans that improve their quality of life. For robotizing these tests, we are designing and developing CLARC, a mobile robot able to help the physician to capture and manage data during the CGA procedures, mainly by autonomously conducting a set of predefined evaluation tests. Built around a shared internal representation of the outer world, the architecture is composed of software modules able to plan and generate a stream of actions, to execute actions emanated from the representation or to update this by including/removing items at different abstraction levels. Percepts, actions and intentions coming from all software modules are grounded within this unique representation. This allows the robot to react to unexpected events and to modify the course of action according to the dynamics of a scenario built around the interaction with the patient. The paper describes the architecture of the system as well as the preliminary user studies and evaluation to gather new user requirements.

A Unified Internal Representation of the Outer World for Social Robotics

Enabling autonomous mobile manipulators to collaborate with people is a challenging research field with a wide range of applications. Collaboration means working with a partner to reach a common goal and it involves performing both, individual and joint actions, with her. Human-robot collaboration requires, at least, two conditions to be efficient: a) a common plan, usually under-defined, for all involved partners; and b) for each partner, the capability to infer the intentions of the other in order to coordinate the common behavior. This is a hard problem for robotics since people can change their minds on their envisaged goal or interrupt a task without delivering legible reasons. Also, collaborative robots should select their actions taking into account human-aware factors such as safety, reliability and comfort. Current robotic cognitive systems are usually limited in this respect as they lack the rich dynamic representations and the flexible human-aware planning capabilities needed to succeed in these collaboration tasks. In this paper, we address this problem by proposing and discussing a deep hybrid representation, DSR, which will be geometrically ordered at several layers of abstraction (deep) and will merge symbolic and geometric information (hybrid). This representation is part of a new agents-based robotics cognitive architecture called CORTEX. The agents that form part of CORTEX are in charge of high-level functionalities, reactive and deliberative, and share this representation among them. They keep it synchronized with the real world through sensor readings, and coherent with the internal domain knowledge by validating each update.