A system of shared autonomous vehicles for Chicago: Understanding the effects of geofencing the service

Gurumurthy, Krishna Murthy; Auld, Joshua; Kockelman, Kara · 2021 · Crossref

DOI: 10.5198/jtlu.2021.1926

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the operational impacts of shared autonomous vehicles (SAVs) in the 20-county Chicago region, specifically focusing on how geofencing service areas affects system performance. Motivated by concerns that widespread SAV adoption could increase congestion and vehicle-miles traveled (VMT) due to empty vehicle miles traveled (eVMT) required for pickups, the authors explore whether restricting SAV operations to specific geographic zones can mitigate these negative externalities. The research aims to determine if spatial constraints, combined with dynamic ridesharing (DRS), can reduce eVMT, lower response times, and ensure equitable access compared to unrestricted regional service. The researchers utilized POLARIS, an agent-based discrete-event transport simulator, to model travel behavior across the Chicago region’s detailed network of 31,000 links. The simulation incorporated modules for destination choice, mode choice, and congestion feedback, alongside a vehicle ownership model predicting reduced personal car reliance. Four scenarios were tested: service restricted to the City of Chicago, the city plus suburban core, the core plus exurban areas, and the entire region without a fence. Each scenario was evaluated under two operational modes: single-occupant travel and DRS. Fleet sizes were calibrated based on resident ratios (1 SAV per 100 residents for mixed-mode scenarios and 1 per 10 residents for SAV-only scenarios), with fares set at 50 cents per mile. Results indicate that geofencing significantly influences SAV efficiency. Service areas with high trip densities, such as the City of Chicago and suburban core, exhibited lower eVMT percentages and more uniform response times compared to sprawling exurban areas or unfenced regions. In the City of Chicago, eVMT was approximately 14%, whereas unfenced operations saw higher eVMT and greater spatial inequity, with peripheral zones experiencing response times exceeding 20 minutes. DRS proved most effective at reducing VMT and eVMT in sprawling regions where trip density was lower, though it had limited impact in dense urban cores where demand was already high. When SAVs replaced all personal auto trips, system-wide VMT increased by up to 2%, highlighting the necessity of maintaining transit systems to prevent congestion. The study concludes that geofencing is a viable policy tool for managing SAV fleets, particularly in reducing eVMT and ensuring equitable service levels. However, the effectiveness of geofencing depends heavily on land-use patterns; dense areas naturally support efficient SAV operations, while sprawling areas require DRS to achieve similar benefits. The findings suggest that fleet sizing must be carefully calibrated to local trip densities rather than maximizing vehicle utilization across a 24-hour period. Ultimately, the research implies that a combination of geofencing, DRS, and preserved transit infrastructure is essential to realizing the potential benefits of SAVs without exacerbating congestion or inequity.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• driverless ads