Pilot Study in Maryland Finds Automated Parking Technology Allowing Vehicles to Park as Close as 10 Centimeters to Adjacent Vehicles Can Lead to a Possible Parking Capacity Expansion of up to 20 Percent.
A Rapid Assessment of Maryland’s Infrastructure Readiness and the Benefits Transit Agencies Can Obtain from Deploying Automated Parking Technology.
Date Posted

Connected and Automated Parking Feasibility – A Pilot Study

Summary Information

Motivated by the fact that Maryland has multiple parking facilities that are at capacity such that commuters spend time searching for parking spaces to a point that can negatively impact their decision to take the commuter rail or rapid transit, this study evaluated a pilot deployment of an Automated Valet Parking (AVP) technology, implemented from March 2019 to June 2021, to test the benefits transit agencies can obtain from deploying automated parking technology. This study is a part of the Transit Innovations Deserving Exploratory Analysis (IDEA) Program, funded by the Federal Transit Administration (FTA) as part of the Transit Cooperative Research Program (TCRP) to support development and testing of innovative solutions for advancing transit practice. The deployment involved operation of low-speed automated vehicles in the parking lots of two Maryland Area Regional Commuter (MARC) train stations in suburban Maryland, equipped with a software kit enabling vehicles to drive and park autonomously within mapped parking lots and garages (equivalent to Society of Automotive Engineers [SAE] Level Four autonomous driving). In addition, four surveys of 164 MARC commuter rail riders were conducted during the pilot study.


Equipped with the software kit, a Connected and Automated Vehicle (CAV) can drop off a passenger at the entrance to the train station and then self-park in a dedicated section of the parking lot. The passenger can then use their smartphone application to summon the CAV to pick them up upon arrival at the station.

For this pilot study, MARC Dorsey station was selected as the initial location, where the MARC Dorsey station was digitally mapped and closed-course testing was performed in the station’s parking facility (Phase One). This phase tested a total of 94 unique scenarios with a total of 980 test runs over four weeks between September 2019 to October 2019. Phase Two began in January 2020 and involved off-peak, mixed traffic testing at the MARC Dorsey station at first but later moved to MARC Odenton station due to COVID-19. Phase Two tested a total of 50 unique scenarios with 780 test runs over four weeks. Phase Three (January 2021 - March 2021) was held entirely at the MARC Odenton station and involved 66 unique scenarios with a total of 1,015 test runs over the four weeks for mixed traffic testing during peak travel hours. Scenarios were designed and conducted for a variety of factors including localization, collision avoidance, and interactions with pedestrians and other road users. Key performance metrics focused on scenario score and success rate (e.g., if the driverless vehicle achieves the pre-defined scenario goal ≥90 percent of the time). Four surveys were distributed to 164 commuters before the deployment and during the three testing phases to understand the commuter opinion of CAVs, gauge how CAVs would potentially impact ridership, and collect user feedback during the testing phases.


  • AVP could enable more spaces in the existing lots as parking space sizes can be decreased with no need to enter or exit the autonomously parked vehicles. Deployment test results indicated that the AVP technology allowed the CAVs to park within 10 centimeters of the adjacent vehicles, translating into an as much as 20 percent increase in the number of vehicles that can be accommodated in the same space, which is a percentage in agreement with previous research.
  • The convenience of an AVP will positively impact commuter behavior and increase ridership by saving the commuters extra effort and time by eliminating the need to park and then walk a long distance to the station from that spot.
  • AVP could allow transit agencies to increase parking revenue and reduce capital costs as a result of the potential parking capacity expansion.
  • Commuter survey revealed that the most respondents would be comfortable using the AVP technology, and that its availability made it more likely that they would choose to ride commuter rail.
  • Nearly 60 percent of the 141 respondents stated that they would utilize AVP technology if it was available at their transit station. 
  • Over 23 percent of the 160 survey respondents stated that they would ride transit more often if the parking lot was less full or if they did not have to park their own car.
  • With the caveat of low sample sizes due to COVID-19, 74 percent of the respondents (17 out of 23) stated that an automated vehicle that picked them up at the station would make them feel safer, and about 50 percent of the respondents (12 out of 23) said they would commute on transit more often if they could more accurately predict how long it would take them to park.
  • Roughly 45 percent of the survey respondents (16 out of 36) stated that seeing AVP technology in operation had increased their interest in automated vehicle technology.
Results Type