EcoRouteX

Inspiration

Last-mile delivery is one of the most carbon-intensive parts of e-commerce logistics. Traditional delivery systems mainly optimize for distance or travel time, while ignoring the carbon emissions produced by different routes. This inspired us to develop EcoRouteX, a framework that combines carbon-aware routing with order consolidation to reduce emissions from last-mile deliveries.

What we learned

Through this project, we learned how to model a delivery network as a weighted graph and optimize routes using distance, travel time, and CO₂ emissions. We also learned how DBSCAN clustering can group geographically and temporally compatible orders into consolidated delivery trips. Most importantly, our experiments showed that order consolidation has a much greater impact on emissions reduction than carbon-aware route selection alone.

How we built it

EcoRouteX uses a two-phase approach:

Order Consolidation: Delivery orders are grouped using DBSCAN based on geographical proximity and delivery time-window compatibility.
Carbon-Aware Routing: The consolidated orders are sequenced and routed using a weighted cost model based on distance, travel time, and CO₂ emissions.

We generated a synthetic 10 km × 10 km urban road network with 576 nodes and evaluated the framework using 50, 150, and 300 orders. We compared EcoRouteX with shortest-distance, fastest-time, and carbon-weighted routing without consolidation. 1

Challenges

One of the main challenges was designing a realistic experimental environment without relying on live OpenStreetMap or OSRM routing services. We therefore generated a synthetic road network with varying congestion and elevation conditions. Another challenge was determining how much of the emissions reduction came from routing versus order consolidation. We addressed this using controlled experiments and statistical analysis.

Results

EcoRouteX achieved 9.7%–56.8% CO₂ reduction compared with the individual-delivery baselines. Our analysis showed that carbon-aware routing alone contributed only about 1.6%–2.2%, while order consolidation produced the majority of the emissions reduction. 2

This project demonstrated that, for the tested scenarios, consolidating multiple deliveries into fewer trips is a more effective lever for reducing last-mile emissions than simply selecting a more carbon-efficient route.