China
Africa
Explore chapters and articles related to this topic
Professor Rodney Brooks' Robots
Published in Junichi Takeno, Self-Aware Robots, 2022
Prof. Rodney Brooks at MIT proposed a new robot design methodology called subsumption architecture. This design methodology may be said to be a parallel processing type as opposed to the conventional serial processing type of robot. The robot’s reaction is quick because the sensors and drive systems are directly coupled via behavioral modules. Many behavioral modules are available and are stacked in layers one above the other from low to high levels.
Control paradigms for mobile robots
Published in Arkapravo Bhaumik, From AI to Robotics, 2018
Other than these facets of design, three more salient features of subsumption architecture are low cost, reduced computational power and re-usability of code. Subsumption is modular and therefore cheaper to build and test and a single layer of behaviour is enough to get the robot running. Such results are most appreciable in walking robots, hexapods and humanoids.
Agent Architectures
Published in Weiming Shen, Douglas H. Norrie, Jean-Paul A. Barthès, Multi-Agent Systems for Concurrent Intelligent Design and Manufacturing, 2019
Weiming Shen, Douglas H. Norrie, Jean-Paul A. Barthès
The subsumption architecture has been used to implement a variety of robots with different abilities. It could also be used to implement AGV (automated guided vehicles) and other relatively simple ‘robotic’ agents in agent-based manufacturing systems.
BASTA: BDI-based architecture of simulated traffic agents
Published in Journal of Information and Telecommunication, 2020
Inga Rüb, Barbara Dunin-Kȩplicz
Basta derives mainly from subsumption architecture (Brooks, 1986): the modelled CAVs are mainly reactive while sophisticated deliberation is performed rarely and serves only to optimize driving. One of the striking similarity is the mechanism of decision-making: a hierarchical structure, which consists of layers responsible for different competencies. In both cases, a vehicle sorts the desires by priority and takes them all into consideration. However, in the subsumption architecture, such a design leads to the problematic complexity of high-level algorithms, which have to ensure that their concepts are compliant with low-level needs. We have simplified the reasoning by reversing the order of competencies: the process starts from high-level algorithms, and the final plan is acknowledged by the most crucial desire. That way, consecutive modifications to a planned commitment are automatically important enough to be applied, and our solution needs no interfaces or communication between the layers.