Society 5.0 and Japan’s Cyber-Physical Architecture
Driven by the national "Society 5.0" framework—which integrates cyberspace modeling with physical-world actuation to address demographic aging, resource constraints, and industrial competitiveness—Japanese research institutes and corporations achieved major technological milestones between 2024 and 2026. Spanning aerospace engineering, high-performance and quantum computing, sovereign artificial intelligence, and bio-hybrid robotics, this strategy reveals a distinct "Japanese Model" of innovation defined by three core pillars: hardware-software hybridization, technological sovereignty, and ultra-precision engineering.
While the 2024–2026 period delivered high-profile breakthroughs—including JAXA's pinpoint SLIM lunar landing, the RIKEN-Fujitsu 256-qubit superconducting quantum computer, the launch of the Zetta-scale FugakuNEXT initiative, and the October 2026 launch of the Martian Moons eXploration (MMX) mission—it also exposed engineering vulnerabilities in commercial lunar descent sensors and the H3 launch vehicle's payload support structure that required rigorous root-cause remediation before H3's June 2026 return to flight. Synthesizing these developments illuminates how Japan is coupling autonomous orbital and robotic actuators with next-generation silicon and quantum architectures.
Aerospace Engineering: SLIM Pinpoint Landing, HAKUTO-R, H3, and Orbital Robotics
In January 2024, JAXA's Smart Lander for Investigating Moon (SLIM) shifted planetary exploration from "landing where it is easy" to "landing where we want," touching down just 55 meters east of its target crater rim.[1] SLIM relied on "Smart Eyes" vision-based navigation: matching real-time camera frames against pre-loaded Kaguya (SELENE) and Lunar Reconnaissance Orbiter maps during orbital descent, then scanning at 50 meters altitude for autonomous obstacle avoidance.[1][2] To settle safely onto a 15-degree slope near Shioli crater, JAXA implemented a two-step impact-dissipating landing sequence (main rear gear contact followed by forward rotation onto auxiliary legs), validating vision navigation for the Martian Moon eXploration (MMX) mission—which subsequently launched aboard H3 Flight 10 on October 20, 2026—despite an engine nozzle loss right before touchdown.[1][2][3][8]
In the private lunar sector, ispace followed its 2023 Mission 1 software altitude-filter failure by launching Mission 2 ("RESILIENCE" lander carrying the "TENACIOUS" micro-rover) in January 2025.[4] During terminal descent on June 6, 2025, RESILIENCE maintained vertical attitude but suffered a Laser Range Finder (LRF) hardware delay that prevented timely braking thrust.[4] In March 2026, ispace consolidated its parallel Japanese Series-3 and U.S. APEX 1.0 lander programs into a unified "ULTRA" lander architecture, signing a July 2026 launch contract with Mitsubishi Heavy Industries to fly Mission 3 aboard an H3 rocket in 2028.[5]
| Mission Parameter | ispace HAKUTO-R Mission 1 (2023) | ispace HAKUTO-R Mission 2 (2025) |
|---|---|---|
| Lander Designation | Series 1 (Unnamed) | Series 1 (RESILIENCE) |
| Failure Phase | Terminal Descent (Hover Exhaustion) | Terminal Descent (Delayed Braking) |
| Root Cause Anomaly | Software Logic: Navigation filter rejected valid altimeter jump over a crater rim as sensor noise. | Hardware Sensor: Laser Range Finder (LRF) failed to deliver valid range data to trigger braking burn.[4] |
| Engineering Takeaway | Upgraded terrain-relative filtering logic. | Redundant rangefinding and unified ULTRA lander platform for Mission 3 (2028 H3 launch).[4][5] |
Meanwhile, Japan's sovereign heavy-lift H3 rocket—powered by first-stage LE-9 expander-bleed thermodynamic cycle engines—succeeded on Flights 2 and 3 in 2024 before suffering an upper-stage setback on Flight 8 (December 22, 2025), when manufacturing delamination in the honeycomb payload support structure expanded during fairing separation, prematurely detaching QZS-5 (Michibiki-5) and damaging second-stage propellant lines prior to its second ignition.[6] Following structural reinforcements and non-destructive inspection upgrades, JAXA returned H3 to flight on June 12, 2026 (Flight 6, H3-30 boosterless configuration) and successfully launched MMX aboard H3 Flight 10 in October 2026.[7][8] Conversely, Japanese orbital robotics excelled: Astroscale's ADRAS-J used model-matching autonomous navigation to approach within 15–50 meters of a non-cooperative tumbling H-IIA upper stage in 2024 ahead of active robotic capture on ADRAS-J2, while Gitai demonstrated TRL-7 dual-arm extravehicular servicing outside the ISS Bishop Airlock and aboard an autonomous 16U servicer satellite.[9][10][11][12][13]
Next-Generation Computing: FugakuNEXT and 256-Qubit Quantum Scaling
In 2025, RIKEN and Fujitsu initiated FugakuNEXT (targeted for ~2030), pivoting from Fugaku's homogeneous CPU design to a heterogeneous CPU-GPU architecture built for Zetta-scale AI.[14][15][16] Targeting over 600 ExaFLOPS (0.6 ZettaFLOPS) in FP8 precision, FugakuNEXT pairs FUJITSU-MONAKA—a 2nm Arm processor featuring 144 cores per socket in a 3D ultra-low-voltage layout—with NVIDIA GPUs linked via coherent NVLink Fusion interconnects to unify physics simulations and trillion-parameter neural network training.[15][16][17]
Concurrently, in April 2025, RIKEN and Fujitsu unveiled a 256-qubit superconducting quantum computer, quadrupling their 2023 64-qubit system.[18][19] By employing a vertical 3D chip-stack packaging and wiring layout inside the dilution refrigerator, engineers resolved the millikelvin microwave cabling and thermal-load bottleneck.[18] Linked directly to HPC clusters for hybrid Variational Quantum Eigensolver (VQE) workflows—and complemented by Toshiba's classical FPGA/GPU Simulated Bifurcation Machine (SBM), which achieves 100x speedups over simulated annealing on combinatorial Ising problems—Japan has operationalized hybrid quantum-classical computing.[20][21]
Sovereign and Evolutionary AI: Fugaku-LLM, Sakana AI, and Self-Driving Labs
To prevent foreign black-box dependency and preserve linguistic sovereignty, Japan prioritized domestic foundation models:
- Fugaku-LLM, Sarashina, and tsuzumi: Released in May 2024, the 13B-parameter Fugaku-LLM was trained on Fugaku's A64FX CPUs via Tofu Interconnect D using Megatron-DeepSpeed, excelling on Japanese MT-Bench benchmarks.[22][23] In late 2025, SoftBank deployed its domestic Sarashina Large Telecom Model, while NTT launched tsuzumi 2, a lightweight LLM for low-power, on-premise medical and financial deployment.[24][25]
- Sakana AI & Closed-Loop Labs: Within Tokyo's AI ecosystem, Sakana AI introduced Evolutionary Model Merge (EMM)—breeding open-source model weights and topologies into EvoLLM-JP and EvoVLM-JP—alongside "The AI Scientist," an autonomous research system whose promise was tempered by a 42% code execution failure rate in 2025 audits.[26][27][28] Meanwhile, NIMS deployed NIMS-OS, linking Bayesian optimization AI with robotic synthesis to discover solid polymer battery electrolytes in one month rather than six years.[29][30]
Advanced Robotics: Bio-Hybrid Actuators, Cybernetic Avatars, and Disaster Response
In February 2025, Professor Shoji Takeuchi's University of Tokyo lab unveiled the Multiple Muscle Tissue Actuator (MuMuTA) in Science Robotics: rolling cultured human myoblasts in collagen gel into thin "sushi-roll" threads to maximize nutrient diffusion and prevent necrosis, then bundling them into linear actuators that powered an 18-cm bio-hybrid robot hand.[31][32] The lab also anchored living skin to robotic faces using V-shaped ligament-like perforations.[33]
Under Moonshot Goal 1, Hiroshi Ishiguro's Cybernetic Avatar initiative and the Kanai "Internet of Brains" project (demonstrated in Dubai in November 2025) use Brain-Machine Interfaces (BMI) so paralyzed operators can control remote avatars.[34][35] Facing a projected shortage of 570,000 eldercare workers by 2040, Waseda University deployed the AIREC humanoid caregiver alongside Fujitsu's aiGait dementia-screening analyzer, while disaster teams fielded quadrupedal UGVs after the Noto earthquake, the rubble-crawling SMURF robot, and Tohoku University's flying Dragon Firefighter.[36][37][38][39]
Conclusion: Hybridization, Sovereignty, and Social Utility
Japan's 2024–2026 trajectory demonstrates a cohesive cyber-physical strategy grounded in hybridization (CPU-GPU Zetta-scale supercomputing, 256-qubit quantum-HPC coupling, and living muscle actuators), sovereignty (domestic LLMs, H3 launch systems, and orbital debris remediation), and social utility—aligning frontier computation and robotics with eldercare, disaster resilience, and lunar and Martian exploration.
Updates… (Nov 16, 2026): Following the December 2025 H3 Flight 8 failure (traced to manufacturing delamination in the honeycomb payload support structure), JAXA successfully returned H3 to flight on June 12, 2026 (Flight 6) and launched the Martian Moons eXploration (MMX) mission aboard H3 Flight 10 on October 20, 2026; meanwhile, ispace consolidated APEX 1.0 into its unified ULTRA lander in March 2026 and contracted a 2028 H3 launch for Mission 3.
Works Cited
- JAXA (2024). Smart Lander for Investigating Moon (SLIM) Results from the Moon Landing. https://global.jaxa.jp/press/2024/01/file/jaxa_doc01_20240125_e.pdf
- JAXA / ISAS (2023). 20 minutes of terror: SLIM will attempt a pinpoint accurate landing on the lunar surface. https://cosmos.isas.jaxa.jp/20-minutes-of-terror-slim-will-attempt-a-pin-point-accurate-landing-on-the-lunar-surface/
- The Planetary Society (2024). SLIM, Japan's precision lunar lander. https://www.planetary.org/space-missions/slim-japans-precision-lunar-lander
- ispace, inc. (2025). ispace Releases Technical Cause Analysis for HAKUTO-R Mission 2. https://www.ispace-inc.com/2025/06/24/ispace-releases-technical-cause-analysis-for-hakuto-r-mission-2/
- ispace, inc. (2026). ispace and Mitsubishi Heavy Industries Agree to Launch Mission 3 ULTRA Lunar Lander Aboard H3 Rocket. https://www.ispace-inc.com/2026/07/29/ispace-and-mitsubishi-heavy-industries-agree-to-launch-mission-3-ultra-lunar-lander-aboard-h3-rocket/
- JAXA (2025). Launch Failure of the 8th H3 Launch Vehicle and Setting up a Special Task Force. https://global.jaxa.jp/press/2025/12/20251222-2_e.html
- JAXA (2026). Launch Result of H3 Launch Vehicle flight No.6, 30 configuration Test Vehicle. https://global.jaxa.jp/press/2026/06/20260612-1_e.html
- JAXA (2026). Launch time of the Martian Moons eXploration (MMX) spacecraft aboard H3 Launch Vehicle Flight No.10. https://global.jaxa.jp/press/2026/10/20261018-1_e.html
- JAXA (2024). CRD2 Phase I / ADRAS-J Update: Fly-Around Observation Images of Space Debris Released. https://global.jaxa.jp/press/2024/07/20240730-1_e.html
- Astroscale (2024). Astroscale's ADRAS-J Achieves Historic 15-Meter Approach to Space Debris. https://www.astroscale.com/news/astroscales-adras-j-achieves-historic-15-meter-approach-to-space-debris
- Astroscale (2024). Astroscale Japan Selected for Phase II of JAXA's Commercial Removal of Debris Demonstration Program. https://www.astroscale.com/en/news/astroscale-japan-selected-for-phase-ii-of-jaxas-commercial-removal-of-debris-demonstration-program
- GITAI (2024). GITAI Completes Fully Successful Technology Demonstration Outside the ISS. https://gitai.tech/2024/03/19/gitai-completes-fully-successful-technology-demonstration-outside-the-iss/
- GITAI (2025). GITAI Successfully Completes Space Demonstration of In-House Developed 16U Satellite. https://gitai.tech/2025/01/27/gitai-successfully-completes-space-demonstration-of-in-house-developed-16u-satellite/
- Fujitsu (2025). Fujitsu awarded contract to design next-generation flagship supercomputer FugakuNEXT. https://www.fujitsu.com/global/about/resources/news/press-releases/2025/0618-01.html
- Matsuoka, S. (2025). The Convergence of HPC, AI, and Quantum: Towards FugakuNEXT “made with Japan”. https://www.hpcuserforum.com/wp-content/uploads/2025/10/Satoshi-Matsuoka_The-Convergence-of-HPC-AI-and-Quantum-FugakuNEXT_Oct-2025-HPC-UF.pdf
- RIKEN (2025). RIKEN launches international initiative with Fujitsu and NVIDIA for "FugakuNEXT" development. https://www.riken.jp/en/news_pubs/news/2025/20250822_1/index.html
- NVIDIA (2025). RIKEN, Japan's Leading Science Institute, Taps Fujitsu and NVIDIA for Next Flagship Supercomputer. https://blogs.nvidia.com/blog/fugakunext/
- Fujitsu & RIKEN (2025). Fujitsu and RIKEN develop world-leading 256-qubit superconducting quantum computer. https://www.fujitsu.com/global/about/resources/news/press-releases/2025/0422-01.html
- RIKEN & Fujitsu (2023). RIKEN and Fujitsu develop 64-qubit quantum computer. https://www.riken.jp/en/news_pubs/news/2023/20231005_2/index.html
- Toshiba Digital Solutions (2025). Quantum-Inspired Optimization Solutions SQBM+ Technologies. https://www.global.toshiba/ww/products-solutions/ai-iot/sbm/sbm-technologies.html
- Toshiba Digital Solutions (2025). Quantum-Inspired Optimization Solutions SQBM+ News and Topics. https://www.global.toshiba/ww/products-solutions/ai-iot/sbm/news.html
- Fujitsu (2024). Release of “Fugaku-LLM” – a large language model trained on the supercomputer “Fugaku”. https://www.fujitsu.com/global/about/resources/news/press-releases/2024/0510-01.html
- Tokyo Institute of Technology (2024). Release of "Fugaku-LLM" — a large language model trained on the supercomputer "Fugaku". https://www.titech.ac.jp/english/news/2024/069223
- SoftBank Corp. (2025). SoftBank Corp.'s Large Telecom Model Evolved into a Domestic AI Model and Launched for Internal Use. https://www.softbank.jp/en/corp/news/press/sbkk/2025/20251029_01/
- NTT Group (2025). NTT's Next-Generation LLM "tsuzumi 2" Now Available. https://group.ntt/en/newsrelease/2025/10/20/251020a.html
- AWS Startups (2024). Letting nature lead: How Sakana AI is transforming model building. https://aws.amazon.com/startups/learn/letting-nature-lead-how-sakana-ai-is-transforming-model-building
- Akiba, T., Shing, M., Tang, Y., Sun, Q., & Ha, D. (2024). Evolutionary Optimization of Model Merging Recipes. arXiv:2403.13187. https://arxiv.org/html/2403.13187v1
- Beel, J., Kanoulas, E., et al. (2025). Evaluating Sakana's AI Scientist for Autonomous Research: Wishful Thinking or an Emerging Reality Towards 'Artificial Research Intelligence' (ARI)? arXiv:2502.14297. https://arxiv.org/html/2502.14297v2
- Tamura, R., Tsuda, K., & Matsuda, S. (2023). NIMS-OS: an automation software to implement a closed loop between artificial intelligence and robotic experiments in materials science. Science and Technology of Advanced Materials: Methods, 3(1). https://doi.org/10.1080/27660400.2023.2232297
- NIMS (2024). Aligning AI and Robotic System for Autonomous Automated Materials Explorations. https://nimsnow.nims.go.jp/en/research/2024no3_r03/
- Ren, X., Morimoto, Y., & Takeuchi, S. (2025). Biohybrid hand actuated by multiple human muscle tissues. Science Robotics, 10(99), eadr5512. https://doi.org/10.1126/scirobotics.adr5512
- EurekAlert! / University of Tokyo (2025). Biohybrid hand gestures with human muscles. https://www.eurekalert.org/news-releases/1072918
- The University of Tokyo (2024). Robots face the future. https://www.u-tokyo.ac.jp/focus/en/press/z0508_00360.html
- Moonshot Goal 1 / Internet of Brains (2025). Joint Research with Dubai Future Labs for Multi-platform Robot/Avatar Control System through BMI Technology. https://brains.link/en/news/4128
- Araya Inc. (2025). "From Mind to Motion: Exploring the Future of Bio-Signal Interfaces for Robots and Avatars" (Dubai, November 26–27, 2025). https://www.araya.org/en/publications/event20251113/
- Digital Watch Observatory (2025). Japan turns to AI and robotics to tackle dementia crisis. https://dig.watch/updates/japan-turns-to-ai-and-robotics-to-tackle-dementia-crisis
- Ghost Robotics (2024). Autonomous Robots for Earthquake Disaster Relief. https://www.ghostrobotics.io/case-studies/autonomous-robots-for-earthquake-disaster-relief
- Horizon Magazine (2024). Robots to the rescue: miniature robots offer new hope for search and rescue operations. https://projects.research-and-innovation.ec.europa.eu/en/horizon-magazine/robots-rescue-miniature-robots-offer-new-hope-search-and-rescue-operations
- Yamauchi, Y., Maezawa, Y., Ambe, Y., Konyo, M., & Tadakuma, K. (2023). Development of a remotely controllable 4 m long aerial-hose-type firefighting robot. Frontiers in Robotics and AI, 10, 1273676. https://doi.org/10.3389/frobt.2023.1273676