The key bridge collapse forced new choices on residents affected by the changing traffic landscape: Do I leave earlier or come home later? Should I switch to transit, telework, or carpooling? Can I cancel trips altogether to avoid the stress and extra cost?
Understanding these behavioral shifts is essential for designing realistic detour plans and congestion-management strategies, for supporting communities that have fewer alternatives, and for planning resilient infrastructure that matches how people travel.
This research area examines travel behavior from three angles:
- Listening to travelers — survey-based evidence of how daily life changed
- Watching the network — what observed traffic patterns reveal about behavioral adaptation
- Testing policy simulations — Evaluating strategies that could ease the burden
I. Survey-Based Insights
The team conducted a post-collapse travel behavior survey of residents in Baltimore City and surrounding counties. Responses were collected both online and in the field, yielding 222 valid responses after data cleaning. Home and work locations showed strong coverage around the Key Bridge and along major detour corridors, meaning most respondents were directly affected by the closure. The survey asked about Key Bridge use before the collapse, perceived traffic conditions afterward, changes in travel time and out-of-pocket costs, rerouting tools and how helpful they felt, and psychological stress and disruption to daily life.
Key themes from the responses:
- Longer, more unpredictable trips — many travelers reported later arrivals to work or appointments, with travel times that varied widely from day to day
- Shifted departure times — a noticeable share moved their morning departures earlier to beat the worst congestion
- Cut-back on non-essential travel — some respondents reduced shopping, social, and recreational trips to avoid stress and extra fuel costs
- Mixed satisfaction with rerouting tools — navigation apps and highway message signs helped some drivers, but others felt the information was “always late” when responding to fluid traffic patterns
- Real psychological toll — respondents described frustration, anxiety, and fatigue, particularly those with caregiving responsibilities or hourly jobs where schedule flexibility is limited
Statistical results from the first and second round of the travel behavior survey. Respondents described moderate to severe disruptions in their work and leisure time as a result of the bridge collapse.
II. Traffic Data Patterns
Travel behavior also leaves fingerprints in the data. Using probe-based speed and travel-time records from August 2023 to February 2025, the team examined how traffic patterns shifted across major corridors before and after the collapse.
Morning peak — partial recovery. The data show only modest increases in delay on key detour routes (I-95, I-895, MD-295) right after the collapse. Over the following fall and winter, morning travel times partially recovered, especially on Fridays — consistent with some commuters adopting telework or flexible hours.
Evening peak — persistent congestion. The PM picture is much more severe. Some routes saw 200–400% increases in travel-time index compared with pre-collapse conditions. Even a year later, PM delays remained high, indicating that people did not substantially reduce or shift many afternoon and evening trips.
This pattern is consistent with what researchers call “travel inertia”: for work trips, some people can adopt telework or stagger their hours, softening the morning peak. But evening trips — shopping, childcare, social activities, errands — offer much less flexibility. Those trips continue to crowd the network even as congestion worsens.
Where the shifts show up geographically:
- Fort McHenry Tunnel (I-95) and Baltimore Harbor Tunnel (I-895) became the main pressure valves for cross-harbor traffic
- Northbound PM segments consistently showed long-lasting slowdowns, even six months to a year post-collapse
- Some segments of I-695 saw reduced volumes or shifting congestion, consistent with drivers abandoning certain routes in favor of others
Combined with the survey, the data paint a coherent picture: people reshuffled their routines where they could, but structural limits in the network meant that evening travel remained heavily constrained — especially for those who couldn’t change where they live or work.
The area of study for traffic simulation testing. Researchers evaluated key detour routes along the I-95, I-695, and I-895 corridors.
III. Simulation and Policy Testing
To move from diagnosis to solutions, the team built a detailed microscopic traffic simulation of the Baltimore network, covering I-695, the Fort McHenry Tunnel (I-95), the Baltimore Harbor Tunnel (I-895), and the connecting local roads that feed these corridors. The model was calibrated with data from 230 traffic detectors and run for four key time windows: pre-collapse, collapse day, immediate aftermath (port closed), and six months later (schools open, bridge still out of service).
The team tested several behavior-influencing strategies:
- Ramp metering near the tunnels. Adding traffic signals on key on-ramps feeding I-95 and I-895 helped smooth merge operations and reduce shockwaves locally, but alone did not fully resolve PM congestion when demand remained very high.
- Toll adjustments and HOV incentives. Scenarios with toll changes and reduced tolls for vehicles with three or more occupants (HOV3+) encouraged more carpooling, shifting some travelers from single-occupant trips into shared vehicles and improving corridor efficiency.
- Demand management and trip reduction. Scenarios where 20% and 50% of trips were removed — through expanded telecommuting, cutting non-essential travel, combining errands, and favoring closer destinations — produced some of the largest improvements in speeds and delays, especially when combined with carpool pricing incentives.
The simulations show that changing when and how much people travel can be just as powerful — and sometimes more practical — than trying to build or widen more lanes during an ongoing disruption.
Takeaways
- Travelers Adapt Within Their Constraints: Morning travelers with flexible jobs can adjust departure times or work remotely. Evening travelers, caregivers, and frontline workers often cannot, so they absorb the worst of the delays.
- Trustworthy Information Matters: Drivers rely heavily on navigation apps and roadside signs, but inconsistent or late updates erode confidence. Clear, reliable messaging about detours, travel times, and transit options can support better individual decisions.
- Policy Can Effectively Guide Behavior: Telework agreements, flexible hours, and carpool incentives can materially reduce peak demand. Targeted pricing (e.g., HOV discounts) and ramp metering can make existing infrastructure work harder without new construction.
- Household Resources Affect Travel Behavior: Households with fewer transportation choices — single-car families, transit-dependent residents, shift workers — have less freedom to adapt and often face higher time and cost burdens.
