Search

Dans cet article, nous présentons une analyse spatiotemporelle des tournées de livraison à domicile d’une entreprise d’électroménagers qui détient sa propre flotte de véhicules. Plusieurs millions d’observations de géolocalisation GPS issues de ces tournées de livraison sont collectées, traitées et assignées au réseau routier. À la suite de ces analyses spatiotemporelles, nous développons des calendriers quotidiens d’indices de congestion en fonction de l’heure. Des ratios de congestion sectoriels sont calculés afin de déterminer les meilleures heures de départ de livraison tout en évitant la congestion routière. La réduction de la durée des trajets a été quantifiée en comparant les meilleures heures de départs par rapport aux heures habituelles. À partir des données de notre partenaire, les analyses démontrent une réduction potentielle de 22 % de la durée des routes de livraison., In this article, we present a spatio-temporal analysis of delivery routes from GPS points of private vehicles. Several thousand GPS geolocation observations obtained from this delivery routes are collected, processed and map matched to the road network. As a result of these spatio-temporal analyzes, we develop daily congestion index calendars as a function of the time of the day. The sectorial congestion ratios are computed and used to determine the best delivery departure times, in order to avoid road congestion. Using data from our industrial partner, our analyzes indicate a potential reduction of 22% of the duration of the delivery routes.

Dans cet article, nous présentons une analyse spatiotemporelle des tournées de livraison à domicile d’une entreprise d’électroménagers qui détient sa propre flotte de véhicules. Plusieurs millions d’observations de géolocalisation GPS issues de ces tournées de livraison sont collectées, traitées et assignées au réseau routier. À la suite de ces analyses spatiotemporelles, nous développons des calendriers quotidiens d’indices de congestion en fonction de l’heure. Des ratios de congestion sectoriels sont calculés afin de déterminer les meilleures heures de départ de livraison tout en évitant la congestion routière. La réduction de la durée des trajets a été quantifiée en comparant les meilleures heures de départs par rapport aux heures habituelles. À partir des données de notre partenaire, les analyses démontrent une réduction potentielle de 22 % de la durée des routes de livraison., In this article, we present a spatio-temporal analysis of delivery routes from GPS points of private vehicles. Several thousand GPS geolocation observations obtained from this delivery routes are collected, processed and map matched to the road network. As a result of these spatio-temporal analyzes, we develop daily congestion index calendars as a function of the time of the day. The sectorial congestion ratios are computed and used to determine the best delivery departure times, in order to avoid road congestion. Using data from our industrial partner, our analyzes indicate a potential reduction of 22% of the duration of the delivery routes.

As the largest contributor to greenhouse gas (GHG) emissions in the transportation sector, road freight transportation is the focus of numerous strategies to tackle increased pollution. One way to reduce emissions is to consider congestion and being able to route traffic around it. In this paper we study time-dependent minimum cost paths under several objectives (TDMCP-SO), in which the objective function comprises GHG emissions, driver and congestion costs. Travel costs are impacted by traffic due to changing congestion levels depending on the time of the day, vehicle types and carried load. We also develop time-dependent lower and upper bounds, which are both accurate and fast to compute. Computational experiments are performed on real-life instances that incorporate the variation of traffic throughout the day, by adapting Dijkstra’s label-setting algorithm according to different cost computation methods. We show that explicitly considering first-in, first-out (FIFO) consistency using time-varying speeds allows the efficient computation of tight time-dependent bounds. Our computational results demonstrate that the TDMCP-SO is more difficult to solve to optimality but the proposed algorithm is shown to be robust and efficient in reducing the total cost even for large instances in an environment of varying speeds, outperforming those based on the link travel time model and on the smoothing method according to each optimization objective, flexible departure times, and different load patterns.

As the largest contributor to greenhouse gas (GHG) emissions in the transportation sector, road freight transportation is the focus of numerous strategies to tackle increased pollution. One way to reduce emissions is to consider congestion and being able to route traffic around it. In this paper we study time-dependent minimum cost paths under several objectives (TDMCP-SO), in which the objective function comprises GHG emissions, driver and congestion costs. Travel costs are impacted by traffic due to changing congestion levels depending on the time of the day, vehicle types and carried load. We also develop time-dependent lower and upper bounds, which are both accurate and fast to compute. Computational experiments are performed on real-life instances that incorporate the variation of traffic throughout the day, by adapting Dijkstra’s label-setting algorithm according to different cost computation methods. We show that explicitly considering first-in, first-out (FIFO) consistency using time-varying speeds allows the efficient computation of tight time-dependent bounds. Our computational results demonstrate that the TDMCP-SO is more difficult to solve to optimality but the proposed algorithm is shown to be robust and efficient in reducing the total cost even for large instances in an environment of varying speeds, outperforming those based on the link travel time model and on the smoothing method according to each optimization objective, flexible departure times, and different load patterns.

Référence bibliographique : Rol, 15692, Appartient à l’ensemble documentaire : Pho20Rol, Image de presse

Référence bibliographique : Rol, 29354, Appartient à l’ensemble documentaire : Pho20Rol

Référence bibliographique : Destailleur Paris, t. 1, 103, Référence bibliographique : Vidéodisque, 7559-7560, Appartient à l’ensemble documentaire : DesDestail, Appartient à l’ensemble documentaire : Est18Rev1