• UA

    • DIRECT: Distributed Intelligence for Resilient Collaborative Robotics in Extreme Environments

    “DIRECT: Distributed Intelligence for Resilient Collaborative Roboyics in Extreme Environments”

    Funding program: HORIZON-MSCA-2025-SE-01-01

    Contract: 101299316

    Project duration: 48 months (01/09/2026 – 31/08/2030)

    Project consortium:

    UNIVERSITY OF WARWICK, UK – Coordinator
    LOUGHBOROUGH UNIVERSITY, UK
    FONDAZIONE ISTITUTO ITALIANO DI TECNOLOGIA, IT
    RUHR-UNIVERSITAET BOCHUM, DE
    KHARKIV NATIONAL UNIVERSITY OF RADIO ELECTRONICS, UA
    Rahal Technology Limited, UK
    RANPLAN GROUP AB, SE
    ONE SOURCE CONSULTORIA INFORMATICA LDA, PT
    ICTFICIAL OY, FI
    CUMUCORE OY, FI
    BEMAGROUP, TR
    MCS DATALABS, DE
    NATIONAL COLLEGE OF IRELAND, IE
    CROWDHELIX LIMITED, IE
    Zhejiang University, CN
    BEIJING JIAOTONG UNIVERSITY, CN
    Sejong University, KR
    NATIONAL UNIVERSITY CORPORATION THEUNIVERSITY OF TOKYO, JP
    UNIVERSITA DEGLI STUDI DI GENOVA, IT

    The DIRECT project, which focuses on robotic collaboration in extreme environments, is multifaceted and multidisciplinary, integrating robotic control systems, artificial intelligence, human-computer interaction, wireless communication, and distributed computing. These domains converge on a common goal: to enable autonomous and collaborative robotic agents to operate effectively in uncertain, infrastructure-degraded, or hazardous scenarios where human presence is risky or impossible.

    In particular, the DIRECT project seeks to address the following core research and design questions:

    • Fundamental operational challenges faced by the robotic fleet in extreme environments;
    • How can distributed machine learning be leveraged to construct semantically rich, shared representations of extreme environments?
    • What are the optimal strategies to orchestrate sensing, computing, and communication resources to support the aforementioned distributed AI—especially when infrastructure is compromised or absent?
    • How can human–robot interaction enhance shared reasoning and mission planning in extreme environments? In post-earthquake scenarios—where AI alone may lack the contextual awareness or ethical judgment to act autonomously—how can human expertise provide crucial insights into scene understanding, objective prioritisation, and real-time adaptation to evolving risks?
    • How can a testing framework integrate communications, model training, and robotic control in a replicated extreme environment?