Browsing by Author "Ferreira, Marta de Faria Rijo"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- Study and modeling of CTT locker operationPublication . Ferreira, Marta de Faria Rijo; Serrador, António João Nunes; Pato, Matilde Pós-de-MinaAbstract The rise of e-commerce, greatly accelerated by the COVID-19 pandemic, has created a need for more efficient and dependable delivery methods. As a result, alternative delivery points, such as parcel lockers, are being explored as effective solutions for distributing e-commerce products. As e-commerce continues to grow, so does the study of last-mile delivery solutions. Examining the social context, distribution, and density of parcel lockers is crucial in highlighting their importance. Previous research has identified the factors that influence the selection of delivery locations, assessed environmental risks and compared delivery methods. Few studies have been done regarding the social climate and population characteristics without surveys and questionnaires. This thesis studies and understands the social environment of parcel locker locations and density. It also suggests a method to optimise locations for installing new lockers. This study confidently analyses locker usage patterns, by studying 750 Portuguese lockers and more than 200,000 parcels, as well as load and turnover rates, by delving into the demographic characteristics of Portuguese parishes, focusing on age, education, and employment status, and Portuguese municipalities, focusing on gross income. Real-world data from a prominent Portuguese parcel locker provider reveals that education and employment status significantly impact the selection of parcel locker locations. A comparison of the top ten municipalities by gross income, number of boxes available and the number of parcels deposited shows that gross income and locker utilization are strongly correlated. Based on the models built, it can be concluded that the optimal algorithm to use is GBM with a complete dataset. By employing this algorithm, it becomes possible to forecast locker usage by examining the correlation between PL usage, population income, and other pertinent factors.