Machine Learning for Cooperative Driving in a Multi-Lane Highway Environment

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    Abstract

    Most of the research in automated driving currently involves using the on-board sensors on the vehicle to collect information regarding surrounding vehicles to maneuver around them. In this paper we discuss how information communicated through vehicular networking can be used for controlling an autonomous vehicle in a multi-lane highway environment. A driving algorithm is designed using deep Q learning, a type of reinforcement learning. In order to train and test driving algorithms, we deploy a simulated traffic system, using SUMO (Simulation of Urban Mobility). The performance of the driving algorithm is tested for perfect knowledge regarding surrounding vehicles. Furthermore, the impact of limited communication range and random packet loss is investigated. Currently the performance of the driving algorithm is far from ideal with the collision ratios being quite high. We propose directions for additional research to improve the performance of the algorithm.

    Original languageEnglish
    Title of host publication2019 Wireless Days, WD 2019
    PublisherIEEE Computer Society
    ISBN (Electronic)9781728101170
    DOIs
    Publication statusPublished - 13 Jun 2019
    Event2019 Wireless Days, WD 2019 - Manchester Metropolitan University Business School, Manchester, United Kingdom
    Duration: 24 Apr 201926 Apr 2019
    https://wirelessdays2019.net/

    Conference

    Conference2019 Wireless Days, WD 2019
    Abbreviated titleWD 2019
    CountryUnited Kingdom
    CityManchester
    Period24/04/1926/04/19
    Internet address

    Fingerprint

    Learning systems
    Reinforcement learning
    Packet loss
    Communication
    Sensors

    Keywords

    • Autonomous driving
    • Cooperative driving
    • Highway environment
    • Q learning
    • Reinforcement Learning
    • SUMO

    Cite this

    @inproceedings{5b9884808603474aa54f05945112a70d,
    title = "Machine Learning for Cooperative Driving in a Multi-Lane Highway Environment",
    abstract = "Most of the research in automated driving currently involves using the on-board sensors on the vehicle to collect information regarding surrounding vehicles to maneuver around them. In this paper we discuss how information communicated through vehicular networking can be used for controlling an autonomous vehicle in a multi-lane highway environment. A driving algorithm is designed using deep Q learning, a type of reinforcement learning. In order to train and test driving algorithms, we deploy a simulated traffic system, using SUMO (Simulation of Urban Mobility). The performance of the driving algorithm is tested for perfect knowledge regarding surrounding vehicles. Furthermore, the impact of limited communication range and random packet loss is investigated. Currently the performance of the driving algorithm is far from ideal with the collision ratios being quite high. We propose directions for additional research to improve the performance of the algorithm.",
    keywords = "Autonomous driving, Cooperative driving, Highway environment, Q learning, Reinforcement Learning, SUMO",
    author = "Aashik Chandramohan and Mannes Poel and Bernd Meijerink and Geert Heijenk",
    year = "2019",
    month = "6",
    day = "13",
    doi = "10.1109/WD.2019.8734192",
    language = "English",
    booktitle = "2019 Wireless Days, WD 2019",
    publisher = "IEEE Computer Society",
    address = "United States",

    }

    Chandramohan, A, Poel, M, Meijerink, B & Heijenk, G 2019, Machine Learning for Cooperative Driving in a Multi-Lane Highway Environment. in 2019 Wireless Days, WD 2019., 8734192, IEEE Computer Society, 2019 Wireless Days, WD 2019, Manchester, United Kingdom, 24/04/19. https://doi.org/10.1109/WD.2019.8734192

    Machine Learning for Cooperative Driving in a Multi-Lane Highway Environment. / Chandramohan, Aashik; Poel, Mannes; Meijerink, Bernd; Heijenk, Geert.

    2019 Wireless Days, WD 2019. IEEE Computer Society, 2019. 8734192.

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    TY - GEN

    T1 - Machine Learning for Cooperative Driving in a Multi-Lane Highway Environment

    AU - Chandramohan, Aashik

    AU - Poel, Mannes

    AU - Meijerink, Bernd

    AU - Heijenk, Geert

    PY - 2019/6/13

    Y1 - 2019/6/13

    N2 - Most of the research in automated driving currently involves using the on-board sensors on the vehicle to collect information regarding surrounding vehicles to maneuver around them. In this paper we discuss how information communicated through vehicular networking can be used for controlling an autonomous vehicle in a multi-lane highway environment. A driving algorithm is designed using deep Q learning, a type of reinforcement learning. In order to train and test driving algorithms, we deploy a simulated traffic system, using SUMO (Simulation of Urban Mobility). The performance of the driving algorithm is tested for perfect knowledge regarding surrounding vehicles. Furthermore, the impact of limited communication range and random packet loss is investigated. Currently the performance of the driving algorithm is far from ideal with the collision ratios being quite high. We propose directions for additional research to improve the performance of the algorithm.

    AB - Most of the research in automated driving currently involves using the on-board sensors on the vehicle to collect information regarding surrounding vehicles to maneuver around them. In this paper we discuss how information communicated through vehicular networking can be used for controlling an autonomous vehicle in a multi-lane highway environment. A driving algorithm is designed using deep Q learning, a type of reinforcement learning. In order to train and test driving algorithms, we deploy a simulated traffic system, using SUMO (Simulation of Urban Mobility). The performance of the driving algorithm is tested for perfect knowledge regarding surrounding vehicles. Furthermore, the impact of limited communication range and random packet loss is investigated. Currently the performance of the driving algorithm is far from ideal with the collision ratios being quite high. We propose directions for additional research to improve the performance of the algorithm.

    KW - Autonomous driving

    KW - Cooperative driving

    KW - Highway environment

    KW - Q learning

    KW - Reinforcement Learning

    KW - SUMO

    UR - http://www.scopus.com/inward/record.url?scp=85068583711&partnerID=8YFLogxK

    U2 - 10.1109/WD.2019.8734192

    DO - 10.1109/WD.2019.8734192

    M3 - Conference contribution

    AN - SCOPUS:85068583711

    BT - 2019 Wireless Days, WD 2019

    PB - IEEE Computer Society

    ER -