In urban logistics, with the continuous growth of eCommerce, last-mile home delivery has become increasingly challenging. It requires a complex planning and scheduling phase to minimise the transport cost incurred by the company, but often results in a failed delivery due to the unavailability of customers. In recent years, the academic world, in parallel with various companies operating in the industry, has encouraged the identification and use of new technologies and delivery models. In this context, this work aims at evaluating the combination of two last-mile innovative solutions, i.e., crowdshipping and mapping customer behaviour. Ordinary people are contracted when needed to deliver parcels to consumers’ designated destinations according to their probability of being available to receive the order. In this dissertation, we introduce and study a variant of the Vehicle Routing Problem in which customers' Availability Profiles and Occasional Drivers are considered (VRPAPOD). The objective is to minimize the total cost incurred by the company while maximising the percentage of successful deliveries. We model the delivery problem as a mixed-integer program and solve it with a branch-and-price algorithm. To analyse the benefit of the combined use of crowd-shipping transportation system and customers availability profiles (APs), we conduct several experiments in a real context, assessing the behaviour of the combination with respect to two benchmarking models: the traditional delivery system (traditional VRP) and the Vehicle Routing Problem with Occasional Drivers, without availability profiles (VRPOD). The computational experiments take place in an area of 16,2 〖km〗^2 in the metropolitan city of Milan, where 100 customers are randomly extracted. The home-attendance profiles of the clients are organised into five clusters and determined through sociological analyses. Through the study of a base case, followed by a sensitivity analysis, it is possible to examine the performance of our proposal in terms of efficiency, obtaining insights into its behaviour. Results prove that the transportation company can achieve important benefits by using the crowd and scheduling deliveries according to clients' APs, which become more significant in case of high drivers availability.
Nella logistica urbana, con la continua crescita dell'eCommerce, la consegna a domicilio dell'ultimo miglio è diventata sempre più impegnativa. Richiede una complessa fase di pianificazione e programmazione per minimizzare i costi di trasporto sostenuti dall'azienda, ma spesso si traduce in una mancata consegna a causa della non disponibilità del cliente. Negli ultimi anni, il mondo accademico, in parallelo con diverse aziende che operano nel settore, ha incoraggiato l'identificazione e l'utilizzo di nuove tecnologie e modelli di consegna. In questo contesto, il presente lavoro mira a valutare la combinazione di due soluzioni innovative dell'ultimo miglio, ovvero il crowdshipping e la mappatura del comportamento del cliente. Persone comuni vengono usate quando necessario per consegnare pacchi a destinazioni designate dai consumatori in base alla loro probabilità di essere disponibili a ricevere l'ordine. In questa tesi, introduciamo e studiamo una variante del Vehicle Routing Problem in cui vengono considerati i profili di disponibilità dei clienti e i driver occasionali (VRPAPOD). L'obiettivo consiste nel minimizzare il costo totale sostenuto dall'azienda massimizzando la percentuale di consegne riuscite. Abbiamo modellizzato il problema delle consegne come un mixed-integer program e risolvendolo con un algoritmo branch-and-price. Per analizzare il beneficio dell'uso combinato del sistema di trasporto crowd-shipping e dei profili di disponibilità dei clienti (AP), conduciamo diversi esperimenti in un ambiente reale, valutando il comportamento della combinazione rispetto a due modelli di riferimento: il sistema di consegna tradizionale (VRP tradizionale) e il Vehicle Routing Problem with Occasional Drivers, senza profili di disponibilità (VRPOD). Gli esperimenti computazionali si svolgono in un'area di 16,2 〖km〗^2 nella città di Milano, dove 100 clienti sono estratti casualmente. I profili di disponibilità sono organizzati in cinque cluster e determinati attraverso analisi sociologiche. Attraverso lo studio di un caso base, seguito da un'analisi di sensibilità, è possibile esaminare le prestazioni della nostra proposta in termini di efficienza, ottenendo informazioni importanti per quanto riguarda il suo comportamento. I risultati mostrano che l'azienda di trasporto può ottenere importanti vantaggi utilizzando la folla e programmando le consegne in base agli APs dei clienti, che diventano più significativi in caso di alta disponibilità di conducenti.
Combining crowdsourcing and mapping customer behaviour in last-mile delivery : analytical model and simulations
Nasti, Andrea;Molinario, Francesco
2019/2020
Abstract
In urban logistics, with the continuous growth of eCommerce, last-mile home delivery has become increasingly challenging. It requires a complex planning and scheduling phase to minimise the transport cost incurred by the company, but often results in a failed delivery due to the unavailability of customers. In recent years, the academic world, in parallel with various companies operating in the industry, has encouraged the identification and use of new technologies and delivery models. In this context, this work aims at evaluating the combination of two last-mile innovative solutions, i.e., crowdshipping and mapping customer behaviour. Ordinary people are contracted when needed to deliver parcels to consumers’ designated destinations according to their probability of being available to receive the order. In this dissertation, we introduce and study a variant of the Vehicle Routing Problem in which customers' Availability Profiles and Occasional Drivers are considered (VRPAPOD). The objective is to minimize the total cost incurred by the company while maximising the percentage of successful deliveries. We model the delivery problem as a mixed-integer program and solve it with a branch-and-price algorithm. To analyse the benefit of the combined use of crowd-shipping transportation system and customers availability profiles (APs), we conduct several experiments in a real context, assessing the behaviour of the combination with respect to two benchmarking models: the traditional delivery system (traditional VRP) and the Vehicle Routing Problem with Occasional Drivers, without availability profiles (VRPOD). The computational experiments take place in an area of 16,2 〖km〗^2 in the metropolitan city of Milan, where 100 customers are randomly extracted. The home-attendance profiles of the clients are organised into five clusters and determined through sociological analyses. Through the study of a base case, followed by a sensitivity analysis, it is possible to examine the performance of our proposal in terms of efficiency, obtaining insights into its behaviour. Results prove that the transportation company can achieve important benefits by using the crowd and scheduling deliveries according to clients' APs, which become more significant in case of high drivers availability.| File | Dimensione | Formato | |
|---|---|---|---|
|
2021_04_Molinario_Nasti.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: testo della tesi
Dimensione
3.91 MB
Formato
Adobe PDF
|
3.91 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/175054