Title: "Applying Biomorphic SLAM to Cloud Robotics Systems"
Time: Wednesday, February 21st, 12pm
Venue: NB 3/57
Autonomous navigation is very active research area in robotics that is searching for alternative methods to provide intelligent ways to robots navigate in unknown environment, which defines the so called simultaneous localization and mapping (SLAM) problem. Outcomes provided by the neuroscience research indicates that hippocampus play a role in individuals’ space navigation when sequences of movements episodes of an individual can be used to construct a map of an unknown environment, which at the end provides a good source of inspiration to design SLAM
solutions (here called Biomorphic SLAM). On the other side, neuroscience can profit from biologically plausible robotic solutions to test the
hypothesis about how the biological structures really work. Therefore, this loop interleaving robotics and neuroscience advancements benefits both fields of knowledge.
From the robotics’ perspective, the gain in using the Biomorphic SLAM becomes even more valuable when multi-robots systems are the target. Nowadays, Cloud-Robotics Systems are the state of the art in multi-robots systems, demanding solutions for a number of problems, in which the cooperative SLAM (C-SLAM) is a highlight. In C-SLAM, all robots perform SLAM in the same environment, but each of them just manages to cover a small portion of that environment. The goal of a C-SLAM solution is to provide a merge of their partial maps so that the system a whole manages to build a complete map.
This talk discusses the application of the Biomorphic SLAM approach in the C-SLAM problem in the scope of Cloud-Robotics Systems. In this context, the state of the collaborative work between UFRGS and RUB will be presented.