The design and implementation of road condition warning system for drivers

Abstract

Intelligent transportation systems are advanced applications that inform vehicle drivers about road conditions. The main purpose of the intelligent transportation systems is to reduce either tangible or intangible loss for the drivers by ensuring the safety of passengers and vehicles. In this study, a system is designed and implemented using wireless sensor networks to inform vehicle drivers about the condition of the road surface. Icing has been chosen as the primary focus of the study since it is considered to be a big threat to road and driver’s safety. The temperature at 10 cm depth of the road, air temperature, relative humidity, air pressure and conductivity values are used as the input data for the prediction of icing on the road surface. The data were previously collected on Raspberry Pi which is a single-board computer and the data were read and processed instantly via k-nearest neighbor algorithm. Using these collected data, the road surface condition is classified as icy, dry, wet or salty-wet. The analyzed results for the road surface condition are presented to the drivers via a mobile application in real time. The drivers are alerted visually and audibly as they approach the coordinates on the road where risky conditions are present.

Determining Priorities for Handling Environmental Road Sections in Palu City (Case Study in West Palu District)

“…The scoring of road conditions is based on the condition variables of each road section. [17] The road condition data used is secondary data from the Palu City Road Infrastructure Basic Data database, which was collected during a survey by the Palu City Public Works Department in mid-2023. For instance, to calculate the scoring criteria for alternative environmental road sections, Jl.…”

Development of black ice prediction model using GIS-based multi-sensor model validation

“…In another study, simple numerical modelling was used to predict black ice. Related studies include road temperature prediction using the integrated model of the Korea Meteorological Administration (KMA) (Park et al, 2014), heat conduction analysis due to air temperature and humidity of road surface (Sass, 1992), salinity and temperature measurement and analysis of road surface (Xu et al, 2017), and development of a black ice estimation algorithm for sensors (Troiano et al, 2010;Teke and Duran, 2019). As in previous studies, the estimation of the amount and location of black ice was insufficient, and there was no model validation.…”

A Distributed Big Data Analytics Architecture for Vehicle Sensor Data

“…Therefore, many studies focus on the prediction of traffic and road conditions using different methods such as Artificial Neural Networks (ANN) [ 22 , 23 ], Long-Short Term Memory (LSTM) Neural Networks [ 24 ], Bayesian networks [ 25 ], and deep learning [ 26 , 27 ] approaches and methods relying on an ensemble of different single predictors of traffic [ 28 ]. Other interesting approaches focus on warning systems, such as the one proposed by Teke and Duran [ 29 ] or dangerous driving events modelling platforms, such as the one proposed by Alvarez-Coello et al [ 30 ]. More specifically, Alvarez-Coello et al engaged the Random Forest (RF) algorithm as well as a Recurrent Neural Network (RNN) in the design of their platform.…”