CV

Education

• B.S. in Aerospace Engineering, Tokyo Institute of Technology, 2015
• M.S. in Mechanical and Aerospace Engineering, Tokyo Institute of Technology, 2017
• PhD in School of Computing, Tokyo Institute of Technology, 2021~

Work experience

• 11/2013 ~ 6/2016: Internship
  • AXELSPACE
• 4/2017 ~ Now
  • Mitsubishi Electric Corporation, Information Technology R&D Center

Code

• tf2rl: a Deep Reinforcement Learning library written in TensorFlow2

Skills

Programming languages

• C/C++ 8years:
  • Embedded softwares for micro satellites, and robots [proj1], [proj2], [proj3]
  • Neural Networks (MLP and CNN) from scratch for onboard satellite imagery classification [proj]
• C# 5years:
  • An autonomous satellite operation system
  • Signal processing to encode/decode data for communications between satellites and ground stations
• python 5years:
  • A Deep Reinforcement Learning library written in TensorFlow2.0 [github]:
  • Prototype developments of machine learning / data analysis algorithms using scikit-learn, TensorFlow, Caffe
• MATLAB/Simulink 2years:
  • Guidance and control for artificial satellites [proj]
• Web development 3years:
  • HTML/CSS/JavaScript/JQuery/MySQL

Robotics / Embedded systems

• ROS 2years:
  • Prototype developments for research (industrial robots, mobile robots, research projects, etc.)
• Mechatronics 3years:
  • Mechanical design using 3D CAD (Inventor, SolidWorks, Fusion360), and 2D CAD (AutoCAD)
  • Mechanical machines (lathe, milling machine, laser processing machine, 3D printer, etc.)
  • Electrical circuit design (MBE, Eagle)
  • Micro computers (PIC, Arduino, mbed, RaspberryPI, Vertex-4)
• Verilog-HDL 1year:
  • Embeded software for micro satellite

Machine learning

• Reinforcement learning
  • Efficient RL using reference path [paper]
  • Model-based RL to solve complex navigation task

Publications

International Conference

1. Kei Ota, Devesh K Jha, Diego Romeres, Jeroen van Baar, Kevin A Smith, Takayuki Semitsu, Tomoaki Oiki, Alan Sullivan, Daniel Nikovski, Joshua B Tenenbaum, “Data-Efficient Learning for Complex and Real-Time Physical Problem Solving using Augmented Simulation”, RA-L + ICRA2021. arXiv.
2. Kei Ota, Devesh K. Jha, Tadashi Onishi, Asako Kanezaki, Yusuke Yoshiyasu, Yoko Sasaki, Toshisada Mariyama, Daniel Nikovski, “Deep Reactive Planning in Dynamic Environments”, CoRL2020. arXiv.
3. Kei Ota, Yoko Sasaki, Devesh K Jha, Yusuke Yoshiyasu, Asako Kanezaki, “Efficient Exploration in Constrained Environments with Goal-Oriented Reference Path”, IROS2020. arXiv.
4. Kei Ota, Tomoaki Oiki, Devesh K. Jha, Toshisada Mariyama, Daniel Nikovski, “Can Increasing Input Dimensionality Improve Deep Reinforcement Learning?”, ICML2020. arXiv.
5. Kei Ota, Devesh K. Jha, Tomoaki Oiki, Mamoru Miura, Takashi Nammoto, Daniel Nikovski, Toshisada Mariyama, “Trajectory Optimization for Unknown Constrained Systems using Reinforcement Learning”, IROS2019. arxiv.
6. Kei Ohta, Takehiko Koike, Yoichi Yatsu, and Saburo Matunaga, “On-board Satellite Imagery Classification using Convolutional Neural Networks,” 31st International Symposium on Space Technology and Science, 2017-n-10, Matsuyama, Japan, June 3-9, 2017.
7. Yuhei Kikuya, Masanori Matsushita, Masaya Koga, Kei Ohta, Yuki Hayashi, Takehiko Koike, Toshiki Ozawa, Yoichi Yatsu, and Saburo Matunaga ,”Fault Tolerant Circuit Design for Low-cost and Multi-Functional Attitude Sensor Using Real-time Image Recognition,” 31st International Symposium on Space Technology and Science, 2017-f-093, Matsuyama, Japan, June 3-9, 2017.
8. Kei Ohta, Masaya Koga, Sota Suzuki, Kazuyoshi Miyasato, Shota Kawajiri, EuGene Kim and Saburo Matunaga, “Proposal and Results of an Automatic Operation System for Nano Satellites Using Multiple Ground Stations”, 30th International Symposium on Space Technology and Science, 2015-j-19, Kobe, Japan, July 6-10, 2015.

See all publications here.

Awards

• 2013 First place in ARLISS 2013
• 2013 First place in 1st AxelSpace Cup
• 2014 First place in 17th RobotGrandPrix
• 2014 First place in 22nd Satellite Design Contest, Idea Award
• 2015 First place in 4th Project Management Award
• 2015 First place in 23rd Satellite Design Contest, Design Award

Projects

Please check portfolio!

Academic Activities

Presentations

• Tutorial talk, “深層強化学習の基礎・応用”，MIRU2021.

Program Committees

• Reviewer for IEEE RA-L, ICRA, IROS, MRVC-2021@ACML, EcoRL2021@NeurIPS.

Teaching

• Instructor
  • Deep Reinforcement Learning Spring Seminar 2021, Autumun Seminar 2021, The University of Tokyo. I taught a course on deep reinforcement learning at The University of Tokyo in Spring 2021. The course takes a broad (deep) RL algorithms including policy gradient, REINFORCE, actor-critic, PPO, DDPG, TD3, SAC. I co-designed the whole course structure and the teaching material for my class (one of six).
• Teachnig Assistant
  • Deep Reinforcement Learning Summer School (RLSS2020), The University of Tokyo. Teaching Assistant for model-based RL class.