A deep reinforcement learning solution to help reduce the cost in waiting time of securing a traffic light for cyclists

Cyclists prefer to use infrastructures that separate them from motorized traffic. Using a traffic light to segregate car and bike flows, with the addition of bike-specific green phases, is a lightweight and cheap solution that can be deployed dynamically to assess the opportunity of a heavier infras...

Full description

Saved in:
Bibliographic Details
Main Authors: Lucas Magnana, Hervé Rivano, Nicolas Chiabaut
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:Journal of Cycling and Micromobility Research
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2950105924000378
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850065405546070016
author Lucas Magnana
Hervé Rivano
Nicolas Chiabaut
author_facet Lucas Magnana
Hervé Rivano
Nicolas Chiabaut
author_sort Lucas Magnana
collection DOAJ
description Cyclists prefer to use infrastructures that separate them from motorized traffic. Using a traffic light to segregate car and bike flows, with the addition of bike-specific green phases, is a lightweight and cheap solution that can be deployed dynamically to assess the opportunity of a heavier infrastructure such as a separate bike lane. To compensate for the increased waiting time induced by these new phases, we introduce in this paper a deep reinforcement learning solution that adapts the green phase cycle of a traffic light to the traffic. Vehicle counter data are used to compare the DRL approach with the actuated traffic light control algorithm over whole days. Results show that DRL achieves better minimization of vehicle waiting time at every hours. Our DRL approach is also robust to moderate changes in bike traffic. The code used for this paper is available at : https://github.com/LucasMagnana/A-DRL-solution-to-help-reduce-the-cost-in-waiting-time-of-securing-a-traffic-light-for-cyclists
format Article
id doaj-art-e341833f5f354305b8d183d31d4b8d0e
institution DOAJ
issn 2950-1059
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series Journal of Cycling and Micromobility Research
spelling doaj-art-e341833f5f354305b8d183d31d4b8d0e2025-08-20T02:49:01ZengElsevierJournal of Cycling and Micromobility Research2950-10592024-12-01210004610.1016/j.jcmr.2024.100046A deep reinforcement learning solution to help reduce the cost in waiting time of securing a traffic light for cyclistsLucas Magnana0Hervé Rivano1Nicolas Chiabaut2CITI, INSA Lyon-Inria, Université de Lyon, Villeurbanne, France; Corresponding author at: 56 bd Niels Bohr, Villeurbanne, France.CITI, INSA Lyon-Inria, Université de Lyon, Villeurbanne, FranceDépartement de la Haute-Savoie, Annecy, FranceCyclists prefer to use infrastructures that separate them from motorized traffic. Using a traffic light to segregate car and bike flows, with the addition of bike-specific green phases, is a lightweight and cheap solution that can be deployed dynamically to assess the opportunity of a heavier infrastructure such as a separate bike lane. To compensate for the increased waiting time induced by these new phases, we introduce in this paper a deep reinforcement learning solution that adapts the green phase cycle of a traffic light to the traffic. Vehicle counter data are used to compare the DRL approach with the actuated traffic light control algorithm over whole days. Results show that DRL achieves better minimization of vehicle waiting time at every hours. Our DRL approach is also robust to moderate changes in bike traffic. The code used for this paper is available at : https://github.com/LucasMagnana/A-DRL-solution-to-help-reduce-the-cost-in-waiting-time-of-securing-a-traffic-light-for-cyclistshttp://www.sciencedirect.com/science/article/pii/S2950105924000378Deep reinforcement learningTraffic lightCyclistsWaiting time3DQNVehicle counts
spellingShingle Lucas Magnana
Hervé Rivano
Nicolas Chiabaut
A deep reinforcement learning solution to help reduce the cost in waiting time of securing a traffic light for cyclists
Journal of Cycling and Micromobility Research
Deep reinforcement learning
Traffic light
Cyclists
Waiting time
3DQN
Vehicle counts
title A deep reinforcement learning solution to help reduce the cost in waiting time of securing a traffic light for cyclists
title_full A deep reinforcement learning solution to help reduce the cost in waiting time of securing a traffic light for cyclists
title_fullStr A deep reinforcement learning solution to help reduce the cost in waiting time of securing a traffic light for cyclists
title_full_unstemmed A deep reinforcement learning solution to help reduce the cost in waiting time of securing a traffic light for cyclists
title_short A deep reinforcement learning solution to help reduce the cost in waiting time of securing a traffic light for cyclists
title_sort deep reinforcement learning solution to help reduce the cost in waiting time of securing a traffic light for cyclists
topic Deep reinforcement learning
Traffic light
Cyclists
Waiting time
3DQN
Vehicle counts
url http://www.sciencedirect.com/science/article/pii/S2950105924000378
work_keys_str_mv AT lucasmagnana adeepreinforcementlearningsolutiontohelpreducethecostinwaitingtimeofsecuringatrafficlightforcyclists
AT herverivano adeepreinforcementlearningsolutiontohelpreducethecostinwaitingtimeofsecuringatrafficlightforcyclists
AT nicolaschiabaut adeepreinforcementlearningsolutiontohelpreducethecostinwaitingtimeofsecuringatrafficlightforcyclists
AT lucasmagnana deepreinforcementlearningsolutiontohelpreducethecostinwaitingtimeofsecuringatrafficlightforcyclists
AT herverivano deepreinforcementlearningsolutiontohelpreducethecostinwaitingtimeofsecuringatrafficlightforcyclists
AT nicolaschiabaut deepreinforcementlearningsolutiontohelpreducethecostinwaitingtimeofsecuringatrafficlightforcyclists