MIT researchers have built a new navigation system that helps drivers find parking instead of just reaching their destination. The system directs users to the parking lot with the best chance of an open spot, considering walk time and success probability. In tests, the approach saved drivers up to 35 minutes compared to waiting for the closest lot.
MIT researchers developed the system to solve a common frustration. Navigation apps tell drivers how long the trip will take but ignore the time spent circling for parking. This leads to late arrivals, worse congestion, and extra emissions. It also makes public transit seem slower than it really is. Cathy Wu, the Class of 1954 Career Development Associate Professor at MIT, supervised the work. Graduate student Cameron Hickert led the study, which appears in Transactions on Intelligent Transportation Systems.
Standard navigation apps send drivers to a destination without considering parking availability. Drivers often arrive only to find no open spots. They then cruise around looking for parking, wasting time and fuel. This extra driving adds to city congestion and pollution. It also hides the true travel time, making it harder for people to choose transit or biking over driving.
READ ALSO: https://modernmechanics24.com/post/top-lithium-metal-starts-ev-batteries/
The MIT system uses a probability-based approach. It looks at all public parking lots near the destination. It calculates drive time to each lot, walk time from each lot to the destination, and the likelihood of finding a spot. The system then identifies the optimal lot with the lowest expected total time for driving, parking, and walking. It even accounts for what happens if the first lot is full, considering nearby alternatives and their success rates.
The method also factors in other drivers competing for the same spots. Someone else might grab the last space, or spillover from other lots could affect success rates. The framework models all these scenarios to give users the best possible guidance. In simulated tests using real traffic data from Seattle, the system cut total travel time by about 60 percent in congested urban settings compared to waiting for a spot. That equals roughly 35 minutes saved for a driver.
The system is not ready for real-world use yet. It needs reliable data on parking availability. Some lots use sensors or gates that track entries and exits, but these are not widespread. The researchers tested crowdsourced data as an alternative. Users could report open spots through an app, or systems could track how many cars circle or leave lots without parking. The team found crowdsourced observations had an error rate of only about 7 percent compared to actual availability, suggesting this approach could work.
WATCH ALSO: https://modernmechanics24.com/post/isro-gaganyaan-crew-parachute-validation/
If deployed, the system could reduce congestion and emissions while saving drivers time. It could also help people make better travel choices by showing realistic drive times that include parking. Future work will test the system with real-time data across entire cities and explore satellite images as another data source. This system directs drivers to the best parking lot rather than the final destination, offering a small change that could have a big impact on urban mobility.
