Perception and Control for Autonomous Navigation in Crowded, Dynamic Environments

Mobile robots are increasingly being used in complex, dynamic, and human-filled environments, such as providing contact-free services during the COVID-19 pandemic, customer service in supermarkets, last-mile package delivery, and transporting both goods and people on roadways. The main challenge in all of these scenarios lies in safely navigating around people, who are inherently difficult to model, and with other robots, who they may not be able to communicate with. Achieving reliable autonomous navigation in these settings involves the integration of a myriad of disciplines, from multi-modal perception to detect and classify individual objects, multi-target tracking to track these objects over time, trajectory prediction to enumerate possible future actions of each object, and (multi-robot) planning and control algorithms that can account for these possible actions and reason about possible reactive coupling among multiple agents.

The aim of this workshop is to bring together researchers from each of these domains to foster discussions about best practices for their integration in order to yield robust navigation systems that function in the most challenging environments. At the workshop attendees will have the opportunity to learn about new results from established researchers in both academia and industry, to hear about preliminary results from young researchers, and to actively engage in open discussions with one another. We hope this workshop will help the community discover new opportunities for collaborations and effective directions of future research.

Topics of interest include (but are not limited to):

  • Perception (multi-modal sensor fusion, object classification and segmentation, etc.)
  • Multi-target tracking (tracking algorithms, trajectory prediction, etc.)
  • Planning and control in dynamic settings (autonomous vehicles, multi-robot planning, game-theoretic methods, etc.)

Submission:

We invite short papers (max 4 pages, excluding citations) in standard RSS paper format, though workshop submissions need not be anonymized. We welcome submissions within the broad areas of perception, tracking, and multi-agent planning. All accepted papers will be introduced in lightning talks, presented as posters, and featured on the workshop website.   Papers may be submitted in PDF format via email to the address: dynamic.perception.control.rss21@gmail.com.

We encourage submissions from young researchers, women, minorities, and members of other underrepresented groups.

Important Dates:

June 20, 2021 - Deadline for paper submissions
July 1, 2021 - Notification of acceptance
July 14, 2021 - Workshop (virtual)

Recordings:

Recordings are available on our YouTube playlist.

Zoom:

Here is a link to the workshop.