Headshot

Incoming Assistant Professor
ME and Texas Robotics
UT Austin

csferrazza@berkeley.edu
Google Scholar
Twitter
Linkedin
Bluesky

I will be hiring PhD students to work on humanoid robotics,
dexterous manipulation, tactile sensing, and robot learning in general.
If you are interested in joining my lab, make sure to apply to UT Austin
and mention my name in your application.

I am an incoming assistant professor in robotics at UT Austin. Currently, I am with Amazon Frontiers AI and Robotics (FAR). Until recently, I was a postdoctoral researcher at UC Berkeley, working with Prof. Pieter Abbeel. My research focuses on advancing humanoid intelligence by incorporating priors, inductive biases, and multi-sensory feedback.

I received my Ph.D. at ETH Zurich in 2021, under the supervision of Prof. Raffaello D'Andrea. During my Ph.D. research, I worked on the design of vision-based, data-driven tactile sensors, and the applications of such sensors to robot control and dexterous manipulation.

My Ph.D. thesis was awarded the ETH Medal, and in 2022, I was selected as a Robotics Science and Systems Pioneer. I am also the recipient of an SNSF Postdoctoral Fellowship, a Rising Star Award and a Best Paper Award at the IEEE International Conference on Soft Robotics, and the ETEL Award.

I received a M.Sc. in Robotics, Systems and Control from ETH Zurich in 2016, and a B.Sc. in Automation Engineering from Politecnico di Milano. During my studies, I completed an exchange programme at Tongji University in Shanghai, where I received a Double Bachelor's Degree in Information Technology Engineering (2016).

Read more...

News

September 2025

  • I will be at CoRL in Seoul and give a talk at KAIST AI. Reach out if you want to chat!
  • I gave a guest lecture at the University of Michigan.
  • Two papers accepted at NeurIPS 2025: BRC and SOMBRL.

July 2025

  • We will be hosting a workshop on Tactile Sensing Toward Robot Dexterity and Intelligence at IROS 2025 next October!

June 2025

May 2025

  • Our new work on highly-parallelized off-policy RL for humanoid robots, FastTD3, is on Arxiv! Check out the social media thread.

Apr 2025

  • I gave invited talks at the CMU Robotics Institute, the University of Washington, and UIUC.
  • MuJoCo Playground has been accepted to RSS 2025!

Mar 2025

  • I gave invited talks at Harvard University, UCLA, ETH Zurich, and Boston University.
  • Two workshop proposals accepted: WCBM@RSS and EXAIT@ICML.

Feb 2025

  • I gave invited talks at UT Austin, Johns Hopkins University, and the Max Planck Institute for Intelligent Systems.
  • I was selected as a Rising Star in Soft Robotics at IEEE RoboSoft 2025!
  • Honored to serve as an Area Chair for CoRL 2025!

Jan 2025

  • I gave an invited talk at Northwestern University.
  • Two papers, FuSe and HOP, have been accepted to ICRA 2025!
  • Two papers, MaxInfoRL and ActSafe, have been accepted to ICLR 2025!
  • Our open-source sim-to-real pipeline for robot learning, MuJoCo Playground, is live! Check out the social media thread.
  • Our work on finetuning VLAs with heterogeneous sensory data, FuSe, is on Arxiv. Check out the social media thread.

Dec 2024

  • Our work on a new class of RL algorithms with directed exploration, MaxInfoRL, is on Arxiv. Check out the social media thread.

Nov 2024

  • We held the CoRL workshop on Whole-body Control and Bimanual Manipulation: Applications in Humanoids and Beyond. See some pictures here!
  • I will also be speaking at the CoRL workshop on Mastering Robot Manipulation in a World of Abundant Data.
  • I will be presenting the Body Transformer paper at CoRL in Munich!
  • Honored to be selected as a finalist for the Embassy of Italy AI Award!

Oct 2024

  • I will be presenting M3L at IROS in Abu Dhabi!

Sep 2024

  • Our new work on hand-object interaction pretraining from videos, HOP, is on Arxiv!

Aug 2024

  • Our new work on an embodiment-aware architecture, Body Transformer, is on Arxiv! Check out the social media thread. We gave a preview of this paper at the RSS EARL Workshop, where it won the Best Poster Award!

Jul 2024

  • I will also be speaking at the RSS workshop on Tactile Sensing for General Purpose Robot Learning.
  • I will be presenting HumanoidBench at RSS in Delft!
Show more...

Select Publications

  1. fem
    MuJoCo Playground: An open-source library for
    GPU-accelerated robot learning and sim-to-real transfer
    ● Best Demo Paper Award at RSS 2025 ●
  2. fem
    HumanoidBench: Simulated Humanoid Benchmark for Whole-Body Locomotion and Manipulation
  3. fem
    Beyond Sight: Finetuning Generalist Robot Policies with Heterogeneous Sensors via Language Grounding
  4. fem
    Body Transformer: Leveraging Robot Embodiment for Policy Learning
  5. Show all

Select PhD Talks

This talk summarizes the vision of my PhD research (shorter version here):


An academic talk on my PhD research can be found here.