Robot Learning
Learning policies from multimodal observations and robot demonstrations.
MSc student · Aalto × KTH
Interested in robot learning and robotic manipulation, with hands-on experience in real-robot data collection, policy deployment, and robotic perception.
01 / CURRENT INTERESTS
Learning policies from multimodal observations and robot demonstrations.
Experimenting with foundation models that connect perception, language, and action.
Bimanual systems, deformable objects, and reliable execution on physical robots.
Data collection, training, system integration, evaluation, and practical debugging.
02 / SELECTED PROJECTS
A selection of hands-on work across robot learning, perception, and industrial vision.
▶WATCH LIVE DEMO ↗
Collected 506 real-robot demonstrations on a TRON2 bimanual robot, full-fine-tuned π0.5, and deployed the policy for physical task execution.
A distributed ROS2 and Zenoh pipeline for multi-LiDAR point clouds and PointPillars detection.
WATCH DEMO ↗
An eight-camera, hardware-triggered acquisition system with a deployed ResNet-18 classifier.
03 / EXPERIENCE
LIMX DYNAMICS · SHENZHEN
Worked on embodied AI and robotics solutions involving manipulation, perception, data collection, and real-robot deployment. Evaluated practical trade-offs among VLA-based methods, conventional vision pipelines, and robot control methods.
ALIBABA GROUP, TAOBAO · HANGZHOU
Contributed to a large-scale low-code platform, working across data delivery pipelines, DSL components, server-side rendering, and platform infrastructure.
04 / EDUCATION
AALTO UNIVERSITY × KTH ROYAL INSTITUTE OF TECHNOLOGY
HEBEI UNIVERSITY OF TECHNOLOGY
05 / CONTACT
Have something in mind?
Let’s talk↗