Postdoctoral Research Associate — AI-Driven Reactive Robotics for Radioisotope Production

The Isotope Research & Production (IP) program at Brookhaven National Laboratory has an opening for a postdoctoral researcher to develop an AI-driven robotic system for the chemical purification steps in radioisotope production. The scientific challenge is not “can a robot move a beaker,” which is largely solved, but the open problem underneath it: fast, reactive fine-motor manipulation of fluids, precision pipetting, and pouring that corrects in real time when liquid begins to slosh or spill; coupled to a frontier vision-language model that plans and supervises the multi-step workflow. The successful candidate will have wide latitude to shape the research direction and will publish contributions to reactive, generalizable robotic manipulation in unstructured laboratory environments. This position sits at the intersection of a mission-driven isotope-production program and frontier machine learning, and has a high level of interaction with an interdisciplinary scientific community spanning robotics, radiochemistry, and computing. Essential Duties and Responsibilities: Design and build a dual-system robotic architecture: a frontier vision-language model for task planning, affordance reasoning, and anomaly/stop-gating, coupled to a fast learned controller for reactive fine-motor manipulation. Develop the fast reactive control layer using an approach matched to your background (e.g. force-feedback MPC with a learned motion prior, latent world models, imitation/vision-language-action policies, or continuous-time neural control). Build the 3D perception stack for fluid-state estimation (fill level, meniscus, vessel pose) through transparent, reflective glassware, using pretrained visual encoders. Develop and validate on a cold bench (water and glassware) before any active material. Publish results in peer-reviewed venues and present at conferences. Required Knowledge, Skills, and Abilities: Ph.D. in mechanical engineering, electrical engineering, computer science, physics, mathematics, statistics/data science, or a closely related technical field Backgrounds in robotics, controls, or machine learning are especially relevant. Strong programming skills in Python and hands-on experience with modern ML frameworks (e.g. PyTorch). Hands-on problem-solving skills and clear, concise written and verbal communication. Demonstrated record of peer-reviewed publication. Ability to work independently and drive an open-ended, high-risk/high-reward research agenda. Preferred Knowledge, Skills, and Abilities: Experience in one or more of: robotic manipulation, reinforcement learning, imitation learning, world models, model-predictive/optimal control, or continuous-time dynamical systems. Demonstrated record of peer-reviewed publication, ideally as a lead (first or primary) author. Experience with collaborative robot arms (e.g. UFactory xArm) and force/torque sensing. Experience with vision-language(-action) models,…