Carmelo Sferrazza
Carmelo (Carlo) Sferrazza is a roboticist and senior applied scientist at Amazon Frontiers AI and Robotics (FAR), where he works on humanoid loco-manipulation and perceptive policies. He is also an incoming assistant professor at the University of Texas at Austin and was previously a postdoctoral researcher at UC Berkeley under Pieter Abbeel. He earned his Ph.D. from ETH Zurich under Raffaello D'Andrea, focusing on vision-based tactile sensing and dexterous manipulation. He is known for his work on humanoid robot learning, including projects such as HumanoidBench, the Body Transformer, and Perceptive Humanoid Parkour.
“Amazon's Fundamental AI Research (FAR) lab; co-affiliation for multiple authors including Justin Kerr, Carmelo Sferrazza, Jitendra Malik, and Angjoo Kanazawa”
Source→“Carmelo Sferrazza... Co-author on MuJoCo Playground... a key figure in Amazon's robotics simulation tooling”
Source→“Co-author Carmelo Sferrazza is affiliated with Amazon FAR, indicating Amazon's interest in tactile sensing for warehouse/fulfillment automation. Contact-rich manipulation (plug insertion, pick-and-place) maps directly to fulfillment center tasks”
Source→“Industry co-author from Amazon's frontier AI division; his participation signals that this line of research has direct industrial interest from a major robotics deployer”
Source→AI-extracted from podcast / newsletter / paper summaries. May contain errors.