|
Jeremy Siburian Hi! I'm Jeremy. I am a 1st-year master's student at The University of Tokyo, advised by Yusuke Iwasawa and Tatsuya Matsushima at the Matsuo-Iwasawa Lab's Robotics Group. Prior to UTokyo, I received my bachelor's degree in Mechanical Engineering at Waseda University, where I was advised by Professor Shigeki Sugano and Alexander Schmitz. During my undergrad, I worked on applying tactile sensing for contact-rich manipulation tasks. I am currently a research intern at OMRON SINIC X, mentored by Masashi Hamaya and Cristian C. Beltran-Hernandez. I was also previously a research intern at Daimler Trucks Asia, where I worked on tactile-based robotic bin picking for industrial assembly tasks. Please feel free to reach out! jeremy.siburian@weblab.t.u-tokyo.ac.jp / CV / Scholar / X / Github |
|
|
|
|
ResearchMy ultimate research goal is develop autonomous robots that can be incorporated in our daily lives. I envision robots that can perform long-horizon decision-making in complex environments, while continuously learn new skills throughout their lifetime. To this end, my current research explores foundation models for planning, data-efficient imitation learning, and multimodal representation learning. |
Publications |
|
Jeremy Siburian*, Keisuke Shirai*, Cristian C. Beltran-Hernandez*, Masashi Hamaya, Michael Görner, Atsushi Hashimoto Preprint (In Submission) Project Page / arXiv TLDR: While recent advances in vision-language models (VLMs) have accelerated the development of language-guided robot planners, their black-box nature often lacks safety guarantees and interpretability crucial for real-world deployment. This paper proposes ViLaIn-TAMP, a hybrid planning framework for enabling verifiable, interpretable, and autonomous robot behaviors. |
|
Jeremy Siburian*, Cristian C. Beltran Hernandez*, Masashi Hamaya SII 2025; ICRA 2024 Cooking Robotics Workshop 🎉 Best Video Award 🎉 OpenReview / YouTube TLDR: Our framework integrates PDDLStream, an existing TAMP framework, with the MoveIt Task Constructor, a multi-stage manipulation planner, to enhance multi-step motion planning for interdependent tasks. We augment our framework with various cooking-related skills, such as object fixturing, force-based tip detection, and slicing using Reinforcement Learning (RL). |
|
Jeremy Siburian, Alexander Schmitz, Tito Pradhono Tomo, Sophon Somlor, Gang Yan, Satoshi Funabashi, Shigeki Sugano 42rd Annual Conference of the Robotics Society of Japan (RSJ) 2024; Undergraduate Thesis Paper TLDR: To be able to perform dexterous manipulation, slip detection is a crucial skill for robot hands. In this study, we perform a comparison of different slip detection algorithms using a distributed tri-axial tactile sensor. We evaluate different sets of tactile features to investigate the importance of distributed 3-axis measurements for slip recognition. |
Projects |
|
Jeremy Siburian 2023, Industry Project with Mitsubishi Fuso (Daimler Trucks Asia) Project Link / Code TLDR: Developed a robotic bin picking system for assembly line deployment using 3D tactile sensors for force control and slip detection. Managed an R&D budget of 1.5 million yen (Approx. $10k USD). |
|
Jeremy Siburian 2023, Personal Project Paper (Unpublished) / Code TLDR: Developed a novel presentation scoring algorithm based on the level of audience engagements using machine learning (ML) and neural networks. |
OutreachI am a product of numerous people (family, friends, professors, mentors) who have helped and supported me along the way. To pay it forward, please feel free to reach out about research, master's applications, life in Japan, or anything in general. |
About Me Personally
• I was born and raised in Indonesia. If you are an Indonesian working on robotics and machine learning, let's connect! |
|
This awesome template was adapted from Jon Barron! |