Seminario - Towards Robot Reasoning in Cluttered, Crowded and Dynamic Environments
you are invited to attend the Industrial Lecture, in collaboration with Bosch (https://www.bosch.com/research/), which will take place tomorrow afternoon at 17:00.
As you know, Bosch is one of the International Industrial leader in our fields.
Possible collaboration with our Department (wrt different topics) is under discussion. I will be very delighted to have you on board for the seminar, as well as to discuss possible collaborations.
The seminar is particularly designed to stimulate our students (both from Computer Science and Mathematics) on current trends and fundamental problems considered by leading international industries. Please, forward this message to any phd/master student that can be interested.
The seminar is, of course, free of charge and open to everyone.
The link to attend the meeting is the following:
Please, find at the end of this email the title and abstract of the seminar. I am sorry for the last minute notification. I hope it is not too late to book a slot in your agenda.
Title: Towards Robot Reasoning in Cluttered, Crowded and Dynamic Environments
Abstract:
Computing safe control policies for wheeled mobile robots navigating in densely cluttered and crowded spaces is a difficult task due to several factors, e.g. perception noise, system models mismatch and high uncertainty of human future behaviors. The latter being influenced not only by other surrounding humans but also by environmental cues. In these settings, classical reactive approaches often result in an overly cautious robot that fails to produce a feasible, safe path in the crowd, or plans a large, sub-optimal, perhaps oscillating, detour to avoid humans in the environment. Moreover a robot unaware of human future movements and of surrounding contextual environmental properties may end up in the so known “freezing" robot problem (i.e. robot blocked by the continuous unpredicted human encounters). A unique solution to fully solve the robot navigation problem in cluttered, crowded and dynamic environments is still far ahead of us. In this talk, starting from a brief analysis of the related work, I will present several robot motion planning approaches to solve the aforementioned issues. In particular I will show how the quest for a safe and efficient robot navigation policy requires the improvements of several planning sub-components, i.e. from increasing of their computational efficiency to the generation of human-aware robot's paths and actions based on predictive planning techniques.