The rapid growth of e-commerce has placed unprecedented pressure on last-mile logistics, now one of the most expensive and environmentally impactful segments of the supply chain. In response, Ferreira and Esperança (2025) investigate how integrating electric vehicles (EVs) and artificial intelligence (AI) can jointly transform urban delivery systems into more efficient, sustainable operations. Their study combines a structured literature review with a real-world case study in Lisbon, offering both conceptual and empirical contributions.
The paper begins by positioning last-mile logistics as a “triple bottom line” challenge involving environmental, economic, and social dimensions. Urban deliveries can account for up to 53% of shipping costs, while conventional delivery fleets significantly contribute to congestion, noise, and emissions. Electric vehicles are increasingly seen as a promising solution due to their lower emissions, quieter operation, and long-term cost advantages. However, EV adoption in logistics faces well-known barriers, including range limitations, limited charging infrastructure, and high upfront investment costs. Artificial intelligence is presented as a complementary technology capable of addressing these limitations through route optimization, predictive energy management, and real-time fleet coordination.
A key contribution of the paper is its synthesis of 78 recent academic studies (2018–2025), which reveals that research on EVs and AI in logistics remains fragmented. Many studies focus on either electrification, routing algorithms, or smart-city technologies in isolation. Few provide integrated frameworks or empirical validation in real operational settings. This gap motivates the authors to propose the ECO.Logística framework, which combines EV deployment, AI optimization, and city consolidation centers (CCCs).
The framework reconfigures urban logistics into a hub-and-spoke model. Large vans transport goods to a consolidation center on the city’s outskirts, after which smaller electric vehicles perform last-mile deliveries. AI plays a central role by optimizing routes using real-time data, scheduling charging cycles, monitoring fleet performance, and improving customer communication. The system also integrates feedback loops to continuously refine operations.
The framework is validated through a case study of a Lisbon logistics provider that gradually implemented AI across its delivery network. Using pre- and post-deployment data, the authors assess performance across key indicators, including delivery time, energy consumption, fleet utilization, customer satisfaction, and emissions. The results are significant: delivery times decreased by 15–20%, energy efficiency improved by 10–25%, fleet utilization rose by 30%, customer satisfaction increased by about 20%, and CO₂ emissions fell by 25–40%.
Despite these promising results, the paper highlights important implementation challenges. High investment costs, limited charging infrastructure, data governance issues, and workforce acceptance remain barriers. The authors also emphasize ethical concerns, including algorithmic bias, workforce impacts, and the risk of prioritizing profitable delivery areas over equitable service coverage.
Overall, the study demonstrates that combining EVs, AI, and consolidation centers can create a scalable and adaptive model for sustainable last-mile delivery. It concludes that future research should focus on long-term pilots, governance frameworks, and cross-sector collaboration to enable widespread adoption.
Source: Ferreira and Esperança (2025)