A drone port 101 is managed using a machine learning model, the port has at least one level with a landing pad 105, a number of rovers 109 which transport goods from a flying drone landed on the pad to one of a number of lockers 111. Training the machine learning model involves creating a series of constraints to be met by a created digital model of a number of drone ports, and creating a series of rewards when particular environment states are met by the digital model. Constraints may include correctness, energy consumption, drone or drone charge level, or time. Environment states may include position or status of the drones or rovers, or the status of the lockers, such as occupied, open, closed, locked and/or unlocked. The rewards may incur a penalty for environment states such as repetitive movements, delivery of goods to the wrong destination, charge level of drones or rovers being below a threshold. Also claimed is a system for managing a fleet of flying drones operating from a number of the drone ports, having a collaborative server 150 running a machine learning model to communicate with, and control, the drones and the rovers.


    Access

    Download


    Export, share and cite



    Title :

    Drone port


    Publication date :

    2022-11-23


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    English



    Drone port

    KOJIMA TORU / MORI SATOSHI / KUNII SHINYA et al. | European Patent Office | 2025

    Free access

    DRONE PORT

    SOGA TAKAYUKI / TAKAHASHI KATSUYUKI / ONO YUKI et al. | European Patent Office | 2021

    Free access

    DRONE PORT

    SOGA TAKAYUKI / TAKAHASHI KATSUYUKI / ONO YUKI et al. | European Patent Office | 2021

    Free access

    DRONE PORT

    MAJOE DENNIS | European Patent Office | 2025

    Free access

    DRONE PORT

    KOJIMA TORU / MORI TOMOHITO / KUNII SHINYA et al. | European Patent Office | 2022

    Free access