SCoCaT was accepted at the Conference on Robot Learning (CoRL 2026) in Austin, Texas.
I work on reinforcement learning for robotic autonomy in space. This includes my current work on Reinforcement Learning and its applications in spcaecraft proximity operations and docking. My main focus is on the model-free and model-based reinforcement learning with safety and operational constraints, large-scale simulation and sim-to-real transfer. Past research efforts have included reinforcement learning applied to legged locomotion on highly deformable terrains. My broader interests span robot learning, spacecraft guidance and control, and autonomous systems that can operate reliably beyond Earth.
I research reinforcement learning for spacecraft autonomy and robot learning, with a focus on safe control and real-world deployment.
More about my work
I work on reinforcement learning for robotic autonomy in space. This includes my current work on Reinforcement Learning and its applications in spcaecraft proximity operations and docking. My main focus is on the model-free and model-based reinforcement learning with safety and operational constraints, large-scale simulation and sim-to-real transfer. Past research efforts have included reinforcement learning applied to legged locomotion on highly deformable terrains. My broader interests span robot learning, spacecraft guidance and control, and autonomous systems that can operate reliably beyond Earth.
Updates
Two spacecraft docking papers were accepted for oral presentation at ESA GNC in Sevilla, Spain and iSpaRo 2026 in Cologne, Germany.
Completed the Space Challenges Program at EnduroSat in Sofia, Bulgaria.
Joined SpaceR at the University of Luxembourg as a doctoral researcher on the Ice2Thrust project.
Graduated from the JEMARO Erasmus Mundus programme at Keio University and the University of Genoa, with a master’s thesis at Ishigami Laboratory: “S3 (Stable Stride on Sand): Enhancing Reinforcement Learning for Optimised Deformable Terrain Locomotion of Quadruped Robots.”
Publications
Selected publicationCoRL 2026
Accepted · Conference on Robot Learning
SCoCaT: Success Conditioned Constrained Reinforcement Learning for Spacecraft Docking
A success-conditioned approach that prevents safety-constrained policies from becoming too conservative to complete terminal navigation tasks, validated across spacecraft simulation and an air-bearing hardware testbed.
All publications and preprints
- 2026
- 2026
PINGU: Extending Air-Bearing Spacecraft Emulators with Open-Source Actuators and Learned Control for Contact-Rich Proximity Operations
arXiv preprint arXiv:2609.23554 · Submitted 20 September 2026
- 2026
Learning Adaptive Multi-Task Guidance, Navigation, and Control via Hypernetworks
arXiv preprint arXiv:2607.24292 · Submitted 27 July 2026
- 2026
Autonomous Spacecraft Docking and Fuel Transfer-Ready Alignment for Sustained On-Orbit Operations via Reinforcement Learning with Constraints
13th International ESA Conference on Guidance, Navigation and Control Systems, Sevilla · Accepted, oral presentation
- 2026
Can Learned Models Enhance Spacecraft Docking and Refueling in a Self-Sustaining Space Economy?
IEEE International Conference on Space Robotics (iSpaRo), Cologne · Accepted, oral presentation
- 2025
Reinforcement Learning based 6-DoF Maneuvers for Microgravity Intravehicular Docking: A Simulation Study with Int-Ball2 in ISS-JEM
AI4OPA Workshop at the International Conference on Space Robotics (iSpaRo), Sendai
- 2021
Development Approaches to Intuitive, SSD and Haptics Integrated and Social HRI Systems for Assisting Space Exploration
IEEE CITISIA, Sydney, Australia
Speaking
Massively GPU-accelerated Constrained Reinforcement Learning for Autonomous Spacecraft Docking and Fuel-Transfer Alignment
Oral presentation · 2026 ESA GNC & ICATT Conference · Sevilla, Spain
Can Learned Models Enhance Spacecraft Docking and Refueling in a Self-Sustaining Space Economy?
Oral presentation · IEEE International Conference on Space Robotics · Cologne, Germany
6-DoF Microgravity Docking with Int-Ball2
Doctoral Workshop on Space Robotics · Malaga, Spain
Bringing Offline Voice Recognition to Automotive Grade Linux
Invited talk · AGL (Automotive Grade Linux) All Member Meeting · Berlin
Experience
Doctoral Researcher
SpaceR, SnT · University of Luxembourg
Constrained reinforcement learning, spacecraft docking, large-scale GPU training and sim-to-real transfer for the EU-funded Ice2Thrust project.
Space Cadet
Space Challenges Program · EnduroSat
Intensive programme covering spacecraft missions, subsystems, satellite operations, space law and the new-space ecosystem.
Graduate Researcher
Ishigami Laboratory (Space Robotics Group) · Keio University, Japan
Reinforcement-learning controllers for quadruped locomotion on deformable granular terrain, including sim-to-real evaluation on Unitree Go1.
Field Robot Operator
Field AI · Japan Operations
Operated legged robots on customer sites and troubleshot autonomous field deployments in unstructured environments.
Open Source Contributor
Google Summer of Code · Automotive Grade Linux
Built an offline voice agent for Automotive Grade Linux using Python, gRPC, GStreamer, Flutter and the Yocto Project.
Network Engineer II
Apple
Developed automation tooling for global production network infrastructure.
Education
Ph.D. in Computer Science
University of Luxembourg
Space robotics and reinforcement learningM.S. in Robotics Engineering
JEMARO Erasmus Mundus Joint Master
Keio University, Japan · University of Genoa, ItalyB.E. in Electronics and Communication Engineering
Thapar Institute of Engineering and Technology
Patiala, IndiaAwards
Invited speaker
Automotive Grade Linux All Member Meeting · The Linux Foundation, Berlin
JEMARO Erasmus+ Consortium Scholarship
Awarded to a selected international cohort
Winner · Smart India Hackathon
Cisco DevNet track · Government of India
Winner · Hack36
MNNIT Allahabad
Winner · HashHacks 3.0
Cluster Innovation Centre · University of Delhi
Thapar Merit Scholarship
Thapar Institute of Engineering and Technology
For research collaborations and other projects in robot learning, spacecraft autonomy or safe reinforcement learning: