Monday 10 March 2025
For years, commuters have been frustrated by bus schedules that are often inaccurate and unreliable. A new study has shed light on a solution to this problem: using artificial intelligence to predict when buses will arrive at their next stop.
Researchers from the University of North Carolina at Charlotte analyzed data from Boston’s public transportation system, including information about bus routes, stops, and departure times. They used this data to train a neural network – a type of AI that can learn patterns in complex data – to predict when buses would arrive at each stop.
The team found that their model was able to accurately predict arrival times for 80% of all bus stops, with an average error of just 77 seconds. This is significantly better than traditional methods, which are often based on simple algorithms and can be inaccurate by several minutes.
But how does the AI work? The researchers used a technique called feature engineering to identify relevant patterns in the data. They looked at factors such as the time of day, the day of the week, and even the weather to help their model make more accurate predictions.
For example, they found that buses tend to run slower during rush hour, when traffic is heavier. They also discovered that certain routes are more prone to delays due to construction or accidents. By taking these factors into account, the AI was able to adjust its predictions accordingly.
The researchers hope that their model will be used in real-world applications to improve public transportation systems around the world. This could include integrating the AI with existing bus schedules and arrival displays, allowing passengers to plan their trips more accurately and reducing frustration on crowded buses.
One potential benefit of this technology is improved traffic flow. If buses are able to arrive at stops closer to their scheduled times, it could reduce congestion on roads and make transportation more efficient overall.
The study’s findings also have implications for other areas of public transportation, such as predicting when trains or subways will arrive. The researchers believe that their model can be adapted to work with data from a variety of transportation systems, making it a valuable tool for cities around the world.
In the future, the team plans to continue refining their model and exploring new ways to apply AI to public transportation. With the potential to improve the daily commute for millions of people, this technology has the potential to make a significant impact on our lives.
Cite this article: “AI-Powered Bus Scheduling: A Solution to Inaccurate and Unreliable Commutes”, The Science Archive, 2025.
Artificial Intelligence, Bus Schedules, Public Transportation, Neural Network, Arrival Times, Feature Engineering, Data Analysis, Traffic Flow, Commute, Prediction
