The design of an IoT based traffic monitoring device using ESP32 microcontroller with mini LiDaR and ultrasonic sensor

A miniaturized version of a Light Detection and Ranging (LiDAR) sensor is available and becoming more affordable, which can be used to detect a fast-moving object by measuring the distance. Herein, we tried to compare TF-Luna LiDAR sensor performance with HC-SR04 ultrasonic sensor for traffic monitoring application. The ESP32 was used as the microcontroller to process the data read by the sensors and send the data to the cloud. The device was implemented to monitor the traffic at certain locations around Institut Teknologi Kalimantan area which the accident may likely to occur. Before conducting the tests, both sensors were calibrated to verify the readings prior to the actual distance measurement. The linear regression fits well with the sensors output and the LiDAR reading is slightly better compared to the ultrasonic related to the average error produced and the R-square values. The traffic data was sent to Google spreadsheet using the Wi-Fi connection, and the delay for sending the data along with the ping checking was observed. From the traffic monitoring, the vehicle count rate from the sensors are fairly accurate to the real count, although the device failed to achieve the perfect counting result. The height of the device for the LiDAR reading can be adjusted to 70 cm to have the most accurate results, while it is difficult for ultrasonic sensor. The Gaussian probability plot was used to model the traffic data. The standard deviation from the three locations produced different results, but the lowest is at the second location according to the LiDAR reading. The mean distance reading data shows that most of the vehicles passed the road at about half of the lane.