Humanoids 2012 Workshop on Developmental Robotics
Can developmental robotics yield human-like cognitive abilities?
Osaka, Japan, November 29th, 2012
|
Announcement: An international symposium on "Constructive Developmental Science" will be held on the day before the workshop. We encourage participation.
November 28, 2012, Osaka, Japan. Free of charge. Link. |
| Instead of trying to produce a programme to simulate the adult mind, why not rather try to produce one which simulates the child's? |
| Alan Turing, 1950 |
The state-of-the-art robots designed for specific tasks exhibit abilities that surpass those of human. However, no robot can pass the motor and cognitive capabilities of a 3 years old child. To create machines that parallel or even pass the motor and cognitive capabilities of humans, the Developmental Robotics field was born as an alternative to previous robot-learning or AI programming approaches. This new approach argued that the continual development of an embodied robotic agent following the development steps of a human is key to achieve motor and cognitive skills that of a human. Consequently research efforts based on this view produced impressive results by focusing on different developmental stages of the embodied "infant robots" that develop through interaction with the environment.
The aim of this one-day workshop is two-fold. First, we will provide an overview of the current state-of-the-art in this field, and remind ourselves about the promises of the developmental robotics and the achievements obtained until now. Our speakers from developmental psychology will also discuss cognitive capabilities of human infants in different stages of their development and the possible mechanisms of acquiring these capabilities.
Second, we will motivate our speakers to comment on (1) how different developmental stages and computational models developed so far can be combined to achieve a coherent model that explains all different developmental stages, and (2) how higher-level cognitive competence can emerge in this developmental progression. The majority of computational models and learning methods developed until now correspond to skills of infants 2 years old or younger. We will try to answer the question what is needed for developmental robotics to make the leap to enable infant robots to acquire the higher-level cognitive abilities, such as complex reasoning, symbolic planning and mental state inference.
Contact:
Emre Ugur, PhD
ATR, NIA Labs.
2-2-2 Hikaridai Seika-cho Soraku-gun
Kyoto 619-0288 Japan
email: emre@atr.jp
phone: +81 774 95 2403