Real-Time Traffic Congestion Prediction Using Predictive Data Analysis

Mohapatra, Ambarish G. · 2024 · Crossref

DOI: 10.63503/j.ijaimd.2024.22

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of real-time traffic congestion prediction in urban environments, aiming to improve traffic management systems by forecasting density before congestion occurs. The research is motivated by the limitations of current reactive traffic control methods, which rely on static signal timings and fail to adapt to dynamic conditions. By leveraging predictive data analysis, the authors seek to develop a system that utilizes real-time sensor inputs and historical data to optimize travel times, reduce fuel consumption, and lower emissions. The primary objective is to compare the performance of a traditional time-series model against a proposed hybrid regression model in terms of accuracy, computational efficiency, and responsiveness. The methodology involves processing real-time traffic data collected from IoT sensors over a seven-day period. The dataset includes vehicle counts, average speeds, and occupancy rates recorded at one-minute intervals. Data preprocessing steps included handling missing values via linear interpolation, reducing noise using a simple moving average filter, and normalizing features to a 0–1 range. Two predictive models were developed and tested: Model 1 is an Auto-Regressive Integrated Moving Average (ARIMA) time-series model that relies on historical traffic patterns; Model 2 is a hybrid regression model that combines multiple linear regression with time-series data, incorporating real-time sensor variables and lagged predictors to capture dynamic changes. Model performance was evaluated using Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) on a separate test set of one week’s data. The results demonstrate that the hybrid regression model outperforms the basic ARIMA time-series model. Visual comparisons of predicted versus actual occupancy rates show that the hybrid model tracks actual traffic conditions more closely, particularly during moderate traffic periods, whereas the ARIMA model exhibited deviations during peak or irregular traffic conditions. The hybrid model’s prediction errors were centered around zero, with most errors falling within a narrow range of -10 to +10, indicating high accuracy. The maximum residual error for the hybrid model was 16.17, and the mean residual error was -0.00, suggesting no substantial bias. The analysis also confirmed an inverse correlation between vehicle count and speed, validating the model’s ability to capture fundamental traffic flow dynamics. The significance of this work lies in its contribution to Intelligent Transportation Systems by providing a more accurate and adaptable method for congestion forecasting. The findings suggest that integrating real-time sensor data with regression techniques offers superior performance compared to historical-only time-series approaches. This hybrid approach enables proactive traffic management, allowing authorities to adjust signal timings and redirect traffic flows in real-time. Consequently, the proposed model supports the development of smarter urban infrastructure, enhancing traffic efficiency and improving the quality of life in densely populated cities.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-25; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• traffic density