RIKEN AIP

RIKEN AIP
Name	AIP
Native name	Advanced Intelligence Project
Established	2018
Location	Tokyo, Japan
Type	Research institute
Director	Masahiro Ono
Parent	RIKEN

RIKEN AIP RIKEN AIP is an advanced artificial intelligence research institute located in Tokyo, Japan, focused on machine learning, robotics, and computational science. It combines interdisciplinary teams drawn from international centers and collaborates with universities, corporations, and national laboratories to pursue applied and foundational AI research. The institute engages with a wide network including industrial partners, academic consortia, and governmental advisory bodies to translate research into technology platforms and social applications.

Overview

The institute operates as an intersection of researchers from institutions such as University of Tokyo, Kyoto University, Osaka University, Tohoku University, Keio University, Waseda University, Tokyo Institute of Technology, Nagoya University, Hokkaido University, Tsukuba University, Yokohama National University, University of Oxford, Massachusetts Institute of Technology, Stanford University, Carnegie Mellon University, University of California, Berkeley, University of Cambridge, Princeton University, Harvard University, California Institute of Technology, University of Toronto, ETH Zurich, École Polytechnique Fédérale de Lausanne, Max Planck Society, CNRS, Imperial College London, National Institute of Informatics, RIKEN Center for Advanced Intelligence Project, NHK, Nissan Motor Company, Toyota Motor Corporation, SoftBank Group, Sony Corporation, NTT Corporation, Hitachi, Ltd., Panasonic Corporation, Fujitsu, NEC Corporation, Microsoft Research, Google DeepMind, IBM Research, Facebook AI Research, OpenAI, Baidu Research, Alibaba DAMO Academy, Tencent AI Lab, Samsung Research, Intel Labs, Arm Ltd., NVIDIA Corporation, AMD, Qualcomm and Oracle Corporation in collaborative projects. Its mission aligns with national strategies and international frameworks including advisories from Ministry of Education, Culture, Sports, Science and Technology (Japan), Ministry of Economy, Trade and Industry (Japan), and engagements with multilateral entities like OECD and G7 dialogues.

History

The institute traces its founding to strategic initiatives connected to research nodes such as AIST and historic laboratories like Bell Labs, Mitsubishi Electric Research Laboratories, Hitachi Central Research Laboratory, and collaborations echoing projects from Human Frontier Science Program, Horizon 2020, DARPA, European Research Council, Japan Science and Technology Agency, National Science Foundation, Wellcome Trust, Simons Foundation, and Gairdner Foundation. Early leadership drew comparisons to figures affiliated with Yoshua Bengio, Geoffrey Hinton, Yann LeCun, Demis Hassabis, Andrew Ng, Fei-Fei Li, Jürgen Schmidhuber, Ian Goodfellow, Daphne Koller, Michael Jordan (computer scientist), Ruslan Salakhutdinov, Shun-ichi Amari, Toru Nishikawa, Masahiro Ono, and institutional models from SRI International, Lawrence Berkeley National Laboratory, Los Alamos National Laboratory, Argonne National Laboratory, Riken Kokubunji Campus and Riken Wako Campus. Major programmatic milestones paralleled international initiatives like AlphaGo, ImageNet Large Scale Visual Recognition Challenge, OpenAI Five, GPT-3, BERT, Transformer (machine learning model), and deployments related to robotics competitions such as DARPA Robotics Challenge and RoboCup.

Research Divisions and Programs

Research divisions span topics with lineage to methodologies and benchmark projects: machine learning groups linked to ImageNet, COCO (dataset), MNIST dataset, CIFAR-10, GLUE (benchmark), SuperGLUE, and generative modeling inspired by GANs, Variational Autoencoder, Diffusion models, Transformer, BERT, GPT-3, AlphaFold, DeepMind AlphaStar, Neural Architecture Search, AutoML, Reinforcement Learning" paradigms associated with Q-learning, Policy Gradient methods, Proximal Policy Optimization, and Monte Carlo Tree Search. Robotics and embodied AI programs map to standards in ROS, OpenAI Gym, MuJoCo, Gazebo (software), KUKA, Boston Dynamics, Honda (company), Toyota Research Institute, iRobot, ABB (company), Fanuc, Yaskawa Electric Corporation, and surgical robotics contexts similar to Da Vinci Surgical System. Applied computing divisions work on bioinformatics interfaces resonant with AlphaFold, BLAST, GenBank, Human Genome Project, and medical collaborations recalling National Center for Global Health and Medicine, Tokyo Medical and Dental University, Keio University School of Medicine, Osaka University Hospital, St. Mary's Hospital, Mayo Clinic, and pharmaceutical partners like Takeda Pharmaceutical Company, Astellas Pharma, Roche, Pfizer, Novartis, Eli Lilly and Company.

Facilities and Collaborations

Facilities include high-performance computing clusters comparable to installations at RIKEN Center for Computational Science, Fugaku, Oak Ridge National Laboratory, Lawrence Livermore National Laboratory, CERN, European Organization for Nuclear Research, Brookhaven National Laboratory, Argonne National Laboratory, and cloud partnerships with Amazon Web Services, Google Cloud Platform, Microsoft Azure, Alibaba Cloud, and IBM Cloud. Experimental labs interface with industrial testbeds from Toyota Motor Corporation, Nissan, Sony, Panasonic Corporation, and prototyping workshops associated with Maker Faire communities. International partnerships have connected AIP to consortia including Global Partnership on Artificial Intelligence, Partnership on AI, The Alan Turing Institute, Vector Institute, Canadian Institute for Advanced Research, Australian Centre for Robotic Vision, KISTI, Institute of Electrical and Electronics Engineers, Association for Computing Machinery, NeurIPS, ICML, CVPR, ACL (conference), IROS, ICRA, AAAI and publisher relationships with Nature (journal), Science (journal), Neural Information Processing Systems, Communications of the ACM, IEEE Transactions on Pattern Analysis and Machine Intelligence.

Education and Outreach

Education efforts coordinate with degree programs at University of Tokyo Graduate School, Kyoto University Graduate School, Osaka University Graduate School, Keio University Graduate School, Waseda University Graduate School, Tokyo Institute of Technology Graduate School, and exchange programs similar to collaborations with MIT Media Lab, Stanford Artificial Intelligence Laboratory, Oxford Robotics Institute, Cambridge Computer Laboratory, École normale supérieure, RIKEN BSI, RIKEN Center for Emergent Matter Science, and summer schools mirroring Deep Learning Summer School, NeurIPS workshops, ICML workshops, and outreach through media outlets like NHK, Asahi Shimbun, Yomiuri Shimbun, The Japan Times, Nature News, Science Magazine, and conferences panels at G20 science meetings.

Notable Achievements and Impact

Notable outputs include publications and software that cite methodologies related to Transformer, BERT, AlphaFold, GANs, and contributions to benchmarks such as ImageNet and GLUE. Collaborative projects have influenced industrial deployments in partners like Toyota Research Institute, Nissan, Sony, SoftBank Robotics, Fujitsu, NEC Corporation, and policy advisories delivered to METI (Japan), MEXT, Ministry of Health, Labour and Welfare (Japan), and international bodies such as OECD and UNESCO. The institute's work engages with ethical frameworks referenced by Partnership on AI, IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems, European Commission AI guidelines, and scholarly discourse in outlets including Nature Machine Intelligence, Science Robotics, Journal of Machine Learning Research, Proceedings of the National Academy of Sciences, and Communications of the ACM.

Category:Research institutes in Japan