Multimodal Trajectory Prediction Model Competitively Recognized Internationally10 July 2025, Taipei, Taiwan – Hon Hai Research Institute (HHRI), an R&D powerhouse of Hon Hai Technology Group (Foxconn) (TWSE: 2317), the world’s largest electronics manufacturer and technology service provider, has been recognized for its competitive work in trajectory prediction in autonomous driving technology. The landmark achievements in ModeSeq, taking top spot in the Waymo Open Dataset Challenge and presenting at CVPR 2025, among the world’s most influential AI and computer vision conferences, gathering top-tier tech firms, research institutions, and academic leaders, highlight HHRI’s growing leadership and technical excellence on the international stage. “ModeSeq empowers autonomous vehicles with more accurate and diverse predictions of traffic participant behaviors,” said Yung-Hui Li, Director of the Artificial Intelligence Research Center at HHRI. “It directly enhances decision-making safety, reduces computational cost, and introduces unique mode-extrapolation capabilities to dynamically adjust the number of predicted behavior modes based on scenario uncertainty.”Figure 1: Illustrates the ModeSeq workflow, showing how the model anticipates multiple possible future trajectories (highlighted by red vehicle icons and arrows). It progressively analyzes the scenario and assigns confidence scores (e.g., 0.2) to each potential path.   HHRI’s Artificial Intelligence Research Center, in collaboration with City University of Hong Kong, on June 13, presented "ModeSeq: Taming Sparse Multimodal Motion Prediction with Sequential Mode Modeling" at CVPR 2025(IEEE/CVF Conference on Computer Vision and Pattern Recognition), where its paper was among only the 22% that were accepted. The multimodal trajectory-prediction technology overcomes the limitations of prior methods by both preserving high performance and delivering diverse potential outcome paths. ModeSeq introduces sequential pattern modeling and employs an Early-Match-Take-All (EMTA) loss function to reinforce multimodal predictions. It encodes scenes using Factorized Transformers and decodes them with a hybrid architecture combining Memory Transformers and dedicated ModeSeq layers. The research team further refined it into Parallel ModeSeq, which claimed victory in the prestigious Waymo Open Dataset (WOD) Challenge – Interaction Prediction Track at the CVPR WAD Workshop. The team’s winning entry surpassed strong competitors from the National University of Singapore, University of British Columbia, Vector Institute for AI, University of Waterloo and Georgia Institute of Technology. Building on their success from last year – where ModeSeq placed second globally in the 2024 CVPR Waymo Motion Prediction Challenge – this year’s Parallel ModeSeq emerged triumphant in the 2025 Interaction Prediction track. Led by Director Li of HHRI’s AI Research Lab, in collaboration with Professor Jianping Wang’s group at City University of Hong Kong and researchers from Carnegie Mellon University, ModeSeq outperforms previous approaches on the Motion Prediction Benchmark—achieving superior mAP and soft-mAP scores while maintaining comparable minADE and minFDE metrics. Figure 2: Director Yung-Hui Li (right) and Researcher Ming-Chien Hsu at CVPR 2025 presenting the latest advances in autonomous driving using ModeSeq.   About Hon Hai Research Institute Founded in 2020 under Hon Hai Technology Group, the institute comprises five research centers and one laboratory. Each unit houses high-tech researchers dedicated to forward-looking studies over a 3–7 year horizon. Their mission is to strengthen long-term innovation and product development to support Foxconn’s transformation toward a “Smart First” future and to bolster the company’s “3+3+3” strategic operating model.

Pioneering a new foundation for quantum cryptography with meta-complexity13 June 2025, Taipei, Taiwan  – Hon Hai Research Institute (HHRI), the research arm of Hon Hai Technology Group (Foxconn) (TWSE: 2317), the world’s largest electronics manufacturer and technology service provider, has achieved a significant breakthrough in quantum computing. Researchers from HHRI’s Quantum Computing Division have demonstrated the possibility of constructing quantum cryptography without relying on traditional one-way functions, using instead a novel theoretical framework known as meta-complexity. This groundbreaking result has been accepted at Crypto, the world’s leading conference in cryptography, highlighting HHRI’s advanced research capabilities and its growing influence in the global quantum technology landscape. Figure 1: Overview of the relationships between key quantum cryptographic primitives and quantum computational complexity. Black lines indicate known results or straightforward inferences; red lines highlight HHRI’s new discoveries.This groundbreaking research, led by Dr. Taiga Hirooka, a researcher at HHRI, in collaboration with Professor Tomoyuki Morimae of Kyoto University and scientific advisor to HHRI, is the first to establish a deep theoretical connection between core tools in quantum cryptography—such as one-way puzzles and quantum random number generators—and a fundamental decision problem known as GapK. The GapK problem centers on determining whether a given piece of information is intrinsically complex or can be succinctly described, offering a novel lens through which to understand the foundations of quantum cryptography.   This discovery not only establishes a new theoretical foundation for quantum cryptography, but also paves the way for the development of next-generation Proofs of Quantumness—a critical step toward constructing secure and verifiable quantum technologies. The work aligns with concurrent research from two international teams and is widely seen as marking a new chapter in the evolution of quantum cryptography.   Crypto (International Cryptology Conference) is one of the most prestigious and longest-running academic conferences in the field of cryptography. Established in 1981, it has become a cornerstone of the discipline—alongside Eurocrypt and Asiacrypt—and is one of the three flagship events organized by the International Association for Cryptologic Research (IACR). Crypto attracts leading researchers and industry pioneers from around the globe and has served as the launchpad for numerous foundational advances in cryptographic theory and practice. Publishing at Crypto is widely regarded as one of the highest honors in the field, representing a hallmark of excellence in theoretical rigor, technical innovation, and lasting impact.   HHRI’s latest breakthrough, now accepted at Crypto, not only enhances Taiwan’s presence in the field of quantum cryptography but also highlights Foxconn’s sustained investment and growing expertise in quantum computing. The Hon Hai Research Institute will continue to advance quantum computing research to drive global technological innovation and industrial progress. Crypto 2025：https://crypto.iacr.org/2025/   About Hon Hai Research Institute The institute, founded in 2020 and part of Hon Hai Technology Group (Foxconn), has five research centers. Each center has an average of 40 high technology R&D professionals, all of whom are focused on the research and development of new technologies, the strengthening of Foxconn’s technology and product innovation pipeline, efforts to support the Group’s transformation from "brawn" to "brains", and the enhancement of the competitiveness of Foxconn’s "3+3+3" strategy.

21 May 2025, Taipei, Taiwan – Foxconn continues to advance its digital health initiatives, unveiling a range of innovations at the 2025 GTC Taipei. Highlights include the AI-powered nursing collaborative robot, digital twin smart hospital development, and contributions to the global open-source medical imaging platform MONAI. These achievements will also be spotlighted during keynote sessions at GTC Taipei, underscoring Foxconn’s capabilities as a comprehensive smart healthcare solutions provider.   Leveraging Technology to Address Nursing Workforce Shortages: AI Collaborative Robot Enters Clinical Practice Amid the challenges of an aging population and critical nursing staff shortages, Foxconn has partnered with Taichung Veterans General Hospital (ranked among Newsweek’s World’s Best Smart Hospitals 2025), Kawasaki Heavy Industries, to co-develop Nurabot, an AI-powered nursing collaborative robot.          Nurabot is powered by Foxconn’s proprietary traditional Chinese large language model, FoxBrain, developed by the Hon Hai Research Institute (HHRI). The model supports TTS (text-to-speech), ASR (automatic speech recognition), and NLP (natural language processing), and is deployed via the Hon Hai Data Center. The robot is powered by Isaac for Healthcare, a physical AI platform built on NVIDIA’s three computers for healthcare robotics. NuraBot uses  the NVIDIA Jetson AGX Orin™system-on-moduleand is trainedusingNVIDIA Isaac Sim™ , a reference robotic simulation application built on NVIDIA Omniverse and running on NVIDIA OVX™ for physical-virtual simulation, enabling autonomous navigation, multimodal perception, and real-time environmental modeling.   In clinical applications, Nurabot performs repetitive tasks such as medication delivery, specimen transport, ward patrols, and patient education—helping reduce nurses’ workload by up to 30%. It also enhances care standardization and precision, enabling nursing staff to focus more on core patient care and clinical decision-making. Currently undergoing field trials at Taichung Veterans General Hospital, Nurabot is expected to be formally integrated into the hospital’s nursing team operations by the end of the year. Foxconn further elaborates on this application in the featured talk titled “AI Humanoid Collaborative Nursing Robots: Shaping the Future of Nursing and Care,” showcasing how AI and robotics help alleviate workforce burdens and reshape healthcare environments.   Launching a Smart Hospital Ecosystem: Digital Twin Deployment for Operational Simulation As part of its smart hospital initiatives, Foxconn is collaborating with several medical institutions under planning and construction, including the Evergreen Branch of Taichung Veterans General Hospital, Baishatun Tung Hospital – Mazu Hospital, and Cardinal Tien Hospital. These collaborations involveconstructingphysically accurate digital twins of smart medical environments with NVIDIA Omniverse™ libraries and technologies.   These digital twins are used across three stages of hospital development: planning and design, workflow simulation, and operational optimization. Even in the early stages of construction and operations, hospitals can simulate AI-driven clinical scenarios, optimize spatial flow, and validate service efficiency and patient experience in advance.   With the integration of NVIDIAVideo Search & Summarization Agent (VSS ), hospitals can analyze real-time video data, automatically identify medical events and anomalies, generate visual summaries, and send real-time alerts. These capabilities empower management teams to make timely, data-driven decisions that enhance care quality and operational performance.   These digital twin environments also serve as key deployment sites for Nurabot, enabling simulation of robot routes, task validation, and training scenarios in virtual space. This accelerates clinical onboarding and cross-site implementation, further optimizing hospital operations and realizing truly patient-centered smart healthcare services.   Contributing to MONAI Open-Source Medical AI Models: Advancing AI Imaging Innovation To advance core AI healthcare technologies, Foxconn has developed CoroSegmentater, a coronary artery segmentation model set to be contributed to the      MONAI Model Zoo as a shared open-source asset for the global medical AI community. Built on 256-slice CTA imaging and powered by MONAI’s Auto3Dseg framework, the model achieves high-precision 3D segmentation of cardiac and coronary structures. CoroSegmentater can be deployed on NVIDIA OVX™ and simulated using Isaac Sim™ to reconstruct heart anatomy and vascular dynamics. It supports various clinical applications, including AI-assisted diagnostics, preoperative planning, and patient communication. The model is being featured in a special session at GTC Taipei, presented by David Niewolny, Sr. Director of Healthcare Business Development at NVIDIA. This contribution reflects Foxconn’s R&D strength in smart healthcare algorithms and its commitment to advancing global medical AI through open collaboration.   Driving Digital Health Globalization and Expanding AI Healthcare Impact Through its collaborationwith NVIDIA and the cross-sector network of HiMEDt (Taiwan Digital Health Alliance), Foxconn is actively promoting the international expansion of Taiwan’s smart healthcare innovations. Moving forward, Foxconn will further enhance its smart hospital AI platforms, robotic applications, and clinical AI model development to transition global smart healthcare from proof-of-concept to large-scale implementation—paving the way toward a truly patient-centered intelligent care ecosystem.