If you’ve ever found yourself driving in a bustling city, you are probably all too familiar with the experience: you plot out your route using a navigation app, and once you reach your destination, the real struggle begins – finding a place to park. This ordeal often leads to significant delays, as you hunt for a parking space and then have to walk from there to your final point. It’s frustrating and contributes to urban congestion and increased emissions.
Unfortunately, most, if not all, navigation systems are designed to drop you off at your destination without any regard for the extra time needed to find parking. This can be a deterrent for those considering mass transit, as they might not realize it could be a faster option.
However, researchers at MIT are working on a solution that could be a game-changer. They’ve developed a system that identifies parking lots with the best balance between location and the chances of finding an open spot. Their unique process directs users to the most suitable parking area instead of the destination itself. In studies using actual traffic data from Seattle, this method showed time savings of up to 66% in heavily congested areas. You could potentially slash about 35 minutes off your travel time compared to waiting for a spot to open up at the nearest car park.
The MIT team’s approach calculates all public parking lots near a destination while considering driving distance, walking distance from the lot to the destination, and the likelihood of finding parking. Perhaps most importantly, the system also prepares for scenarios where you reach an ideal parking lot but find no available spaces, considering the proximity and success probability of other nearby lots.
Cameron Hickert, MIT graduate student and lead author of the research paper, gives an insight, “Our framework can handle cases where it might be smarter to try several nearby lots with slightly lower success probabilities rather than hoping for an opening at the higher-probability lot.”
The system also accounts for the actions of other drivers that could impact parking success. Future data could come from several sources, including crowdsourced information or tracking vehicles circling for parking. With advancements, autonomous vehicles might even report open spots they pass by. “Capturing this information, even through simple app interactions, could be vital for informed decision-making,” Hickert states.
In tests with Seattle traffic data, simulating different urban and suburban scenarios, the method developed by the MIT team reduced travel time by about 60% compared to waiting for a spot and by 20% compared to continually driving to the next closest lot. The prospect of using crowdsourced parking data showed promise with an error rate of only 7% as compared to real-time availability. This suggests that it could effectively gather parking probability data.
Going forward, the team will conduct broader studies that will utilize real-time route information throughout the entire city and explore additional data sources like satellite images to estimate possible emissions reductions. “Transportation systems are complex and hard to change, but small improvements can significantly impact decision-making, congestion, and emissions,” says Cathy Wu, senior author of the research.
This research was supported by Cintra, the MIT Energy Initiative, and the National Science Foundation. For further information, you can visit the original news article here or access their full study in Transactions on Intelligent Transportation Systems here.
This website uses cookies.