The Robotics for Society conference will bring together some of the world's leading researchers in robotics to convene in Vancouver, Canada. Part of an ongoing series of conferences held by the Vancouver Society for Cognitive Systems, this year's conference will focus on the interactions between robotics and the societies in which they are embedded. The program includes invited talks as well as student poster presentations. The small scale and intimacy of this conference will provide an inviting setting for researchers from traditionally disparate disciplines to exchange ideas.

Please pre-register online (below), so we can order appropriate quantities of coffee and snacks. Thanks!

Robotics and Everyday Life

Computers, in all their forms, have become pervasive in modern life, and have increased our productivity and our sense of connectedness to family and friends. Today robots are just starting to creep into our daily lives. As of now they are essentially navigation machines which do their main task as a by product of how they move around in the world. As the world's demographics change to a more aged population there is an opportunity for robotics to enter our lives in more pervasive ways and to increase productivity for all.

The key requirements for our robots are (1) the object recognition capabilities of two year old child, (2) the language capabilities of a four year old child, (3) the manual dexterity of a six year old child, and (4) the social understanding of an eight year old child. These are the challenges for robotics research.

Humanoid Robotic Perspective to Neuroscience

In this talk we take the humanoid robotics perspective to Neuroscience in the studies of the brain. We will present a constructive approach in the exploration of brain-like mechanisms. We place our investigations within real-world contexts, as humans do.

Three essential aspects motivate our approach: 1) In engineering - engineers can gain a great deal of understanding through the studies of biological systems, which can provide guiding principles for developing sophisticated and robust artificial systems; 2) scientifically - building a human-like machine and the reproduction of human-like behaviours can in turn teach us more about how humans deal with the world, and the plausible mechanisms involved; 3) society - in turn we will gain genuine understanding that moves us towards the development of systems that can better interact with people/environment.

Talking Robot Mimicking Human Speech Production

Most of speech synthesis is based on a speech acoustic model and its computer simulation. A mechanical speech synthesizer, however, has potential for development of a more realistic talking head as well as bringing new insights about the physical process of speech production.

We developed a 19 degree-of-freedom (DOF) anthropomorphic talking robot WT-6 (Waseda Talker No. 6) mimicking human speech production. It consists of 1-DOF lung, 4=DOF vocal chords and 13 DOF articulators: 7-DOF tongue, 5-DOF lips, 1-DOF jaw and 1-DOF velum. Three dimensional vocal tract shapes are produced by 3-D articulator structures, and vowel and consonant sounds are produced by directly reproducing the aero-acoustic process of speech production.

In this talk, we describe the robot mechanism with focus on the soft tissue articulators: the vocal cords, tongue, and lips; and show demonstration videos of the robotÕs speech behavior. We also discuss the speech mimicking robot control with auditory and tactile feedback.

A Social Robot in the Wild World; Practices in Therapeutic and Pedagogical Applications

We built social robots, Infanoid and Keepon, as examples of implementation of the design principles that make robots capable of embodied interaction with children. After testing their communicative functionality in experimental settings, we proceeded to longitudinal field observations of a group of children with developmental disorders and a group of typically-developing preschool children interacting with Keepon.

From these field studies, we learned that the children changed their ontological understanding of the robot, and so their ways of interactions, as the interaction unfolded. The children, including those with autism, spontaneously engaged not only in dyadic interaction with the robot, but also in triadic interaction among children and careers, where the robot functioned as a pivot of the interpersonal communication. We conclude that the robot's structural and functional simplicity derived from the children's various spontaneous behaviours, whose communicative meanings would be given and enriched by pedagogical and therapeutic environments.

Space for Interaction with a Robot

Currently we have mobile robots that can map their environs both indoors and outdoors, and then reliably recover their metric position while moving around in their space. These arise from solutions to the well-known simultaneous localization and mapping problem (SLAM). But robots cannot yet tell us where they are, nor easily take our directions.

I will describe how we are planning to bridge the gap between dense accurate geometric information and the semantics of the space we share with robots. I will show how to supply a world description to the robot, learned from examples or adopted from ontologies that characterize prototypical spaces. The relations between places and robots support our interactions with them and enable robots to work with us.

You, Robot, Do No Harm!

As we design, build and deploy robots in society, we are confronted with a host of ethical and social issues, some old in new guises, others brand new. Before we can answer any of the pressing ethical questions, we need to consider the presuppositions that they impose on our robot design approaches. Making ethical choices about any of our mutual interactions with robots presupposes that we are able to foresee the possible future effects of that interaction (or inaction).

This presupposition puts strong requirements on the design space for the models we use to represent robot architectures. We must use transparent models of robot structure and function. We need to be able to determine if a robot will satisfy constraints on its future actions. We need languages for modeling robot dynamics (where we take 'dynamics' quite abstractly) and languages for writing constraint-based specifications. Moreover, we need techniques for determining if a robot does (or is likely to) satisfy our specifications. These criteria allow us to evaluate frameworks for robot design.

Our Constraint-Based Agent model satisfies those criteria. The robot and its environment are modeled symmetrically as, possibly hybrid, dynamical systems in Constraint Nets. We emphasize the important special case where the robot controller is a hierarchy of online constraint solvers.

Computational Motor Control, Humanoid Robotics, and their Societal Relevance

Computational motor control is a research field that attempts to understand the neural control of primate movement in terms of mathematical theories of action and perception. Humanoid robotics aims at creating a new kind of robotic systems that can share human living space and infrastructure, with the goal to entertain and assist humans in their daily lives. Many research goals are shared between these two fields.

This talk will describe our work towards understanding human and humanoid motor skill generation and learning using an interdisciplinary approach. At the highest level, we are interested in interactive skill acquisition, for instance using imitation learning. At a lower level, we need to address the basic principles of decomposing human movement into smaller motor primitives, and how to learn such primitives and sequencing of primitives. At the lowest level, dexterous motor control and motor learning needs to be addressed in order to understand the compliant and fault tolerant strategies of human movement and their realization in robotics.

We have accompanied technical research in these areas with behavioral and neuro-imaging studies, and realized our results in humanoid robotics implementation, where a humanoid robot accomplished various complex manipulation and imitation skills. We will also outline how this line of research has important implications for the new wave of societally relevant assistive robotics, e.g., as needed in elder care, robot rehabilitation and physical therapy, intelligent prosthetics, etc.

Assault and Batteries: On the utility of robot aggression, competition and violence

For robots, as for animals, aggression, competition and violence can be useful. In the context of robots, "useful" means increasing the value of the robots to their owners. In the first part of this talk I will describe my laboratory's work in using aggressive behaviour to improve the overall efficiency and utility of groups of robots. Second, I survey some examples of robots with competitive or aggressive behaviour, and consider their impact on the field and on wider society. Finally, I consider the current and future value of robots designed to perform physical violence, and the peculiar ethics thereof.