Max Robotics — Global Service Robot Directory + US Certification

About

Grecia Salazar is a leading researcher at the intersection of robotics, machine learning, and artificial intelligence, whose work is fundamentally reshaping how robots learn and operate in the real world. She is best known for her pioneering contributions to the development of large-scale, generalist robotic models. As the lead author of the landmark "RT-1: Robotics Transformer for Real-World Control at Scale" (512 citations), Salazar demonstrated how a single, transformer-based model could be trained on diverse, task-agnostic data to control a robot across hundreds of varied tasks, achieving unprecedented levels of real-world generalization. She then extended this paradigm with "RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control" (267 citations), a groundbreaking work that showed how internet-scale vision-language models could be fine-tuned to directly output robot actions, enabling emergent semantic reasoning—allowing robots to perform tasks they were never explicitly trained on. Her work on "Q-Transformer" (16 citations) further advanced the field by scaling offline reinforcement learning for multi-task policies. With over 800 combined citations for her core works, Salazar is a key architect of the modern foundation model approach to robotics, pushing the boundaries toward robots that can understand, reason, and act in the open world.