Call for Papers

The RSS 2024 Workshop on Structural Priors as Inductive Biases for Learning Robot Dynamics invites experts in robotics, dynamics, and control, learning with priors, and data-driven systems to present their research to the community.

Topics that intersect with this workshop's research objectives (see also the home page) include but are not limited to:

• Physics-Informed Neural Networks (PINNs) incorporating knowledge about the kinematics and dynamics of robots (e.g., Lagrangian Neural Networks)

• Techniques for deriving reduced-order models enforcing a structure in the latent dynamics (Autoencoders, Dynamic Mode Decomposition, Koopman Operator Theory, Spectral Submanifolds, etc.)

• Latent dynamical models exhibiting a specific structure, such as linear/structured state-space models, Neural ODEs (NODEs), or Recurrent Neural Networks (RNNs).

• Control-oriented model learning for joint system identification and stabilizing controller synthesis

• Leveraging structured, learning-based models (see previous bullet points) for control of robots (e.g., MPC, traditional model-based controllers such as PD + Feedforward, etc.)

Contributions should make some novel algorithmic, theoretical, benchmarking, or perspective advancement. Additionally, contributions may be on-going, under review, or recently published work.

Accepted papers and supplementary material will be made available on this workshop website both before the workshop date and subsequently for future visitors unless otherwise requested by the authors. Sharing papers in this way does not constitute formal proceedings, i.e., this workshop is a non-archival venue that will not restrict later renditions of the work from being published in archival conferences or journals.

Submission Guidelines

The submitted papers should be no more than 4 pages; with no limit on references or supplementary material. Submissions should follow the RSS 2024 paper format (see link). 

Submission Website: https://openreview.net/group?id=roboticsfoundation.org/RSS/2024/Workshop/Priors4Robots

The workshop features a best paper award with a $300 monetary prize. 

There are two streams for workshop contributions:

Poster presentations: You can present posters during the coffee breaks and interact with the workshop audience on your research topic. Furthermore, you might receive valuable feedback on your research project. We'd like for you to submit a one-page (+n pages for references) abstract describing an exciting and innovative research idea.  We also invite submissions that only include preliminary results (e.g., toy examples) for the poster session.  Poster submissions will be assessed based on their expected relevance and contribution to the topics discussed during the workshop.

Spotlight talks: We offer the exciting and unique option to present your research as a 7-minute presentation (+3 minutes for Q&A) to the workshop audience. Submissions aiming for spotlight talks will be peer-reviewed by at least two reviewers, and we expect the extended abstracts (max. 4+n pages) to include substantial algorithmic, theoretical, or benchmarking results that are ideally verified on relevant testbeds (e.g., leveraging priors for learning robot dynamics).

Furthermore, all accepted spotlight talks will compete for the best paper award. If you are interested in giving a spotlight talk at the workshop, please choose the corresponding option when submitting your contribution.

Important Dates

• Submission Deadline: June 1, 2024

• Author Notification: June 20, 2024

• Camera Ready Submission: June 30, 2024

• Workshop: July 15, 2024