Today in Robotics — Monday, November 9, 2026
CoRL season is clearly upon us, with NVIDIA, MIT, and Boston Dynamics all leaning hard into one question: can robots learn meaning, not just motion? Today's lineup pairs fresh takes on simulation and agentic data pipelines with a wheeled Atlas experiment and some decidedly practical warehouse plumbing. Comprehension and dexterity, it seems, are the themes of the day.
In today’s roundup
- Boston Dynamics adds wheels to Atlas humanoid to test whole-body control — The Robot Report
By bolting wheels onto Atlas and blending classical control with whole-body reinforcement learning, Boston Dynamics is probing how dynamic locomotion scales beyond legs alone. - NVIDIA launches GR00T 2 and Isaac Lab 3.0 at CoRL — The Robot Report
NVIDIA's GR00T 2 and Isaac Lab 3.0 arrive at CoRL with a roster of partners already building on them, signaling how much the stack is consolidating around shared tooling. - MIT’s smart system helps robots understand what humans mean, not just what they say — Interesting Engineering
MIT's system aims to close the gap between what people say and what they actually want, a perennial headache for anyone who's ever instructed a literal-minded machine. - Packsize and Swisslog partner to improve warehouse throughput — Automated Warehouse
Packsize and Swisslog pair right-sized packaging with warehouse automation, a reminder that cutting waste and boosting throughput often comes down to the boxes, not the robots. - Can Agentic Robotics Solve Robotics’ Data Scaling Problem? — IEEE Spectrum Robotics
IEEE Spectrum questions the industry's big bet that enough human-demonstration data will crack robotics' hardest problems, and whether agentic approaches offer a shortcut. - FLEX approach helps robots reason beyond literal commands — TechXplore Robotics
MIT's FLEX platform teaches robots to read intent beyond literal commands, chasing the kind of intuitive understanding humans take for granted. - AI agents test and improve robot code in virtual worlds before deployment — TechXplore Robotics
Borrowing the agentic playbook that reshaped software, AI agents now test and refine robot code in virtual worlds before anything touches hardware. - Practice Makes Precise: At CoRL, NVIDIA and Partners Reveal How Robots Rehearse Delicate Work in Simulation — NVIDIA Robotics Blog
Simulation graduates from teaching robots to walk toward finer tasks like aligning gears and connecting cables, where millimeters of precision actually matter.
Read this roundup on RobotCafe →
The Daily Grind · RobotCafe · robotcafe.com
Don't miss what's next. Subscribe to The Daily Grind: