Eco-innovation diffusion in retailers’ last-mile delivery: an Agent-Based Model approach

The transportation field represents cities’ major source of air pollution, causing around 20% of the total GHG emissions at the European level. Among others, road transport contributes the highest share to overall transport emissions. In this framework, cities are starting to introduce zones with low- and zero-emissions, restricting access to Internal Combustion Engine vehicles. We decided to investigate the so-called last-mile delivery and its transition to more sustainable means of freight transport in cities. With this aim, we performed an analysis of the variation in simulated local emissions, varying different transport system parameters, in order to evaluate the possible outcome of different green policies. We used an Agent-Based Model network representation, and we simulated the diffusion of eco-innovation among a retailers’ network, based on real-world data obtained from the Limited Traffic Zone (LTZ) of the centre of Turin, in Italy. Our research shows how combined policies, simultaneously encouraging load efficiency and increasing penetration of ecological vehicles, produce the greatest emission reductions. Success in the diffusion of transport innovation is highly influenced by the network topology, by nodes’ distribution, and by the threshold for adoption. Innovation adoption reaches up to 90% in Small World networks with a 15% threshold, regardless of the ecological node centrality; it drops to ∼70% at 30% threshold, and it falls below 60% at 45% threshold, where only Free Scale and Erdős–Rényi topologies with central ecologic nodes maintain viability, highlighting a steep decline in robustness as adoption thresholds rise.