I’m a robotics and machine learning engineer in San Francisco.
I’ve trained robots to fold cloth and deployed reinforcement learning models in production.
Robot learning · IROS 2022
Training a visual feedback policy in simulation and transferring it to a real robot. The project includes camera latency measurements, a shared C++ controller, and cloth dynamics randomization.
First-author paper. IROS Best Paper Award finalist.
Evaluation · IJRR 2026
Contributed to the ICRA 2024 Cloth Competition’s evaluation service: SAM-assisted segmentation, depth-based cloth coverage measurement, and dataset/server integration.
Co-author of the resulting benchmark and dataset paper.
Computer vision · Stray Robots
Co-founded Stray Robots and co-developed annotation tools combining RGB-D capture, SLAM, 3D reconstruction, and label projection.
Equal-contribution co-author of the 2021 preprint.
The International Journal of Robotics Research
Co-author. Contributed to the competition’s evaluation tooling.
IROS
First author. Best Paper, ABB Best Student Paper, and Best RoboCup Paper finalist.
Reinforcement learning, policy deployment, and inference optimization.
Automated image and video generation, ad optimization, and ML product integration.
Computer vision tools for creating training data from RGB-D scans.
Reinforcement learning and sim-to-real transfer for dynamic cloth manipulation.
Aalto University · MSc in Technology, majoring in Machine Learning, Data Science and Artificial Intelligence. Graduated with honors.
Also a co-founder and board member of Junction, the international hackathon.
I’m interested in developing and evaluating robot learning models through real-world experiments.
julius@juliushietala.com ↗U.S. permanent resident