Dataset Information

Bootstrapping Virtual Bipedal Walkers with Robotics Scaffolded Learning.

ABSTRACT: We reach walking optimality from a very early age by using natural supports, which can be the hands of our parents, chairs, and training wheels, and bootstrap a new knowledge from the recently acquired one. The idea behind bootstrapping is to use the previously acquired knowledge from simpler tasks to accelerate the learning of more complicated ones. In this paper, we propose a scaffolded learning method from an evolutionary perspective, where a biped creature achieves stable and independent bipedal walking while exploiting the natural scaffold of its changing morphology to create a third limb. The novelty of this work is speeding up the learning process with an artificially recreated scaffolded learning. We compare three conditions of scaffolded learning (free, time-constrained, and performance-based scaffolded learning) to reach bipedalism, and we prove that a performance-based scaffold, which is designed by the walking velocity obtained, is the most conducive to bootstrap the learning of bipedal walking. The scope of this work is not to study bipedal locomotion but to investigate the contribution from scaffolded learning to a faster learning process. Beyond a pedagogical experiment, this work presents a powerful tool to accelerate the learning of complex tasks in the Robotics field.

SUBMITTER: Zhu J

PROVIDER: S-EPMC8456033 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Nowadays, robotics applications requiring the execution of complex tasks in real-world scenarios are still facing many challenges related to highly unstructured and dynamic environments in domains such as emergency response and search and rescue where robots have to operate for prolonged periods trading off computational performance with increased power autonomy and vice versa. In particular, there is a crucial need for robots capable of adapting to such settings while at the same time providing robustness and extended power autonomy. A possible approach to overcome the conflicting demand of a computational performing system with the need for long power autonomy is represented by cloud robotics, which can boost the computational capabilities of the robot while reducing the energy consumption by exploiting the offload of resources to the cloud. Nevertheless, the communication constraint due to limited bandwidth, latency, and connectivity, typical of field robotics, makes cloud-enabled robotics solutions challenging to deploy in real-world applications. In this context, we designed and realized the XBot2D software architecture, which provides a hybrid cloud manager capable of dynamically and seamlessly allocating robotics skills to perform a distributed computation based on the current network condition and the required latency, and computational/energy resources of the robot in use. The proposed framework leverage on the two dimensions, i.e., 2D (local and cloud), in a transparent way for the user, providing support for Real-Time (RT) skills execution on the local robot, as well as machine learning and A.I. resources on the cloud with the possibility to automatically relocate the above based on the required performances and communication quality. XBot2D implementation and its functionalities are presented and validated in realistic tasks involving the CENTAURO robot and the Amazon Web Service Elastic Computing Cloud (AWS EC2) infrastructure with different network conditions.

Dataset Information

Bootstrapping Virtual Bipedal Walkers with Robotics Scaffolded Learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets