Stretch Community News - April/May 2024
Welcome to the Hello Robot monthly community update!
Thank you to all the Stretch users and enthusiasts who visited us at ICRA in Yokohama last month! Read on for Stretch assisting with bathing, Sim2Real audio navigation, NeRF-based change detection, Aloha 2 gripper support, open-source SPOC, and more! If you’d like to see your work featured in a future newsletter, please let us know!
Many individuals with physical disabilities need assistance with bathing, a challenging task for assistive robots. Researchers from Carnegie Mellon University have developed a new soft manipulator with capacitive sensing that significantly outperforms the state-of-the-art on limb cleaning effectiveness, while maintaining patient safety and comfort.
In the news: Hello Robot was honored to receive the IEEE Spectrum Technology in the Service of Society award for 2024! Stretch was also prominently featured in an article in the latest edition of MIT Technology Review about the future of useful household robots.
New work from the University of Texas-Austin evaluates the Sim2Real acoustic gap for navigation tasks based on audio-visual signals rather than vision alone, and develops a new strategy for transferring policies for these tasks from simulation to real robots.
“Change detection” is an important concept for robots used in patrol, security, or inspection tasks. Lawrence Technological University researchers have created a new system utilizing NeRF-based image synthesis and CLIP models to allow natural-language user queries that define what type of change is interesting in the environment.
Now you can build and integrate the open-source Aloha 2 gripper with Stretch 3! Adapter part files and software instructions are available on the Hello Robot user forum.
The Allen Institute has made the code and data for their SPOC project publicly available! SPOC is an agent that can proficiently navigate, explore, and manipulate objects in both simulation and in the real world using only RGB sensors, trained by imitating shortest-path experts in simulation.
