Popularity-based Alternative Routing

TL;DR

Polaris introduces a popularity-aware alternative routing framework that uses multiple $K_{\text{road}}$ layers to penalize highly used road edges, thereby reducing congestion and CO$_2$ emissions without requiring centralized coordination. By computing several layers of road popularity from sampled origin-destination trips and applying multi-layer edge-weight penalization, Polaris achieves diverse routes that avoid overusing popular roads. Experimental results across three Italian cities show Polaris lowers the share of highly popular edges and CO$_2$ emissions (up to $23.57\%$ vs baselines) and performs comparably to a coordinated approach, with favorable execution times, highlighting its practical utility for urban traffic management.

Abstract

Alternative routing is crucial to minimize the environmental impact of urban transportation while enhancing road network efficiency and reducing traffic congestion. Existing methods neglect information about road popularity, possibly leading to unintended consequences such as increasing emissions and congestion. This paper introduces Polaris, an alternative routing algorithm that exploits road popularity to optimize traffic distribution and reduce CO2 emissions. Polaris leverages the novel concept of K-road layers, which mitigates the feedback loop effect where redirecting vehicles to less popular roads could increase their popularity in the future. We conduct experiments in three cities to evaluate Polaris against state-of-the-art alternative routing algorithms. Our results demonstrate that Polaris significantly reduces the overuse of highly popular road edges and traversed regulated intersections, showcasing its ability to generate efficient routes and distribute traffic more evenly. Furthermore, Polaris achieves substantial CO2 reductions, outperforming existing alternative routing strategies. Finally, we compare Polaris to an algorithm that coordinates vehicles centrally to distribute them more evenly on the road network. Our findings reveal that Polaris performs comparably well, even with much less information, highlighting its potential as an efficient and sustainable solution for urban traffic management.

Figures (6)

• Figure 1: Comparison of three alternative routes generated by KMD (in red), three alternative routes generated by Polaris (in blue), and the fastest path (in black) between two roads in Milan. The width of the roads is proportional to their popularity, computed as their $K_{\text{road}}$. KMD routes differ significantly from the fastest path but use much more popular roads (56%). In contrast, Polaris minimizes the use of popular road edges (only 7.6%), while still ensuring diverse route options.
• Figure 2: Spatial distribution of road popularity, computed as $K_{\text{road}}$, in Florence (a), Milan (b), and Rome (c). We categorise road edges into Low (cyan), Medium (orange) and High (black) popularity using an equal-sized logarithmic binning on the $K_{\text{road}}$ distribution of each city. Note how, in all cities, there are many highly popular road edges (black road edges).
• Figure 3: Number of vehicles in the simulation versus the number of SUMO teleports for Florence, Milan, and Rome. Simulations are conducted based on the mobility demand of the city and using the fastest route. The red dashed line denotes the elbow point of the curve.
• Figure 4: Comparison of Polaris (black bar) with the baselines in Florence, Milan, and Rome regarding the usage of highly popular edges (a-c, in %), CO2 emissions (d-f, in tons), and traversed regulated intersections (g-i, in %). We show the average and the standard deviation across five simulations.
• Figure 5: Relationship between Polaris parameter $m$ (number of $K_{\text{r}oad}$ layers) and percentage of highly popular edges (a-c), total CO2 emissions (d-f), and traversed regulated intersections (g-i) for three values of $v$ (number of trips used to compute $K_{\text{r}oad}$ values).