Helen Oleynikova

Helen Oleynikova has had a diverse career through academia and industry. She finished her bachelors in robotics in the US at Olin College of Engineering, then worked at Google on Street View for 2 years, and chose to return to robotics by doing a Masters, and then a PhD in Robotics at the Autonomous Systems Lab at ETH Zurich. Her PhD focused on safety and collision avoidance for drones: unifying fast, onlinevolumetric mapping with online replanning to best exploit the structure of the maps. Afterwards, she worked on the DARPA Subterranean challenge as a post-doc at the Autonomous Systems Lab. Returning to industry, she then worked as a Senior Scientist at the Microsoft Mixed Reality and AI Lab: investigating how we can use Mixed Reality to interact with robots, and what role mapping an co-localization can play in that. Afterwards, she joined the Nvidia Isaac 3D Perception team, working on rapidly GPU-accelerating volumetric mapping for a variety of robotics applications: from warehouse AMRs to mobile manipulators. She spent the last 2.5 years back in academica as a Senior Researcher at the Autonomous Systems Lab, supervising a large variety of topics between LiDAR-inertial odometry, robust state estimation, and mobile manipulation, and led EU project on infrastructure maintenance and inspection with both mobile manipulators and drones. Her most recent research interests focus on safety in online learning: how can residual learning and online reinforcement learning be used to compensate for unmodeled errors in the real-world? How can we combine classical priors with learned methods to increase sample efficiency? And finally, how can we make this safe and fast enough to run on robots in real-time? She also has some new career updates in store.