LAUSR

Online grocery shopping is an emerging market. It caused visible growth in last mile delivery services, which has resulted in concerns about its external effects (e.g., pollution, land use, congestion). The study outlined in this paper proposes a grocery delivery concept where goods are transported by train to the customer’s nearest station, and then the consumer either picks up the groceries (i.e., click and collect) or the goods are delivered to the customer’s home, or to a locker. The focus of this paper is primarily based on the last mile delivery part of the supply chain (i.e., from the train station to the customer). The land use efficiency and emissions of each delivery concept have been evaluated based on the time-area concept and the Handbook Emission Factors for Road Transport (HBEFA 4.1), respectively. This large-scale simulation considers every household in Switzerland with various levels of demand and supermarket network densities. Two machine learning techniques (i.e., random forest and decision tree) have been used to categorise all neighbourhoods within Switzerland based on the best delivery method in terms of emissions and land efficiency. The results show that, depending on the scenarios (e.g., 10 or 100% of the households taking part), home delivery can be better for 76–89% of these communities when compared with a click and collect option based on their land use efficiency.