Autonomous Vehicle Speed Regulation Via Traffic Sign Recognition Models

“Autonomous Vehicle Speed Regulation Via Traffic Sign Recognition Models.”

NIKITHA T S1, BRUNDA S2, AKSHITHA R3

1Nikitha T S, Information Science and Engineering, RR Institute of Technology

2Brunda S, Information Science and Engineering, RR Institute of Technology

3Akshitha R, Information Science and Engineering, RR Institute of Technology

Abstract - The intelligent transportation systems have come a long way in a very short time and have consequently created great chances of integrating aboard the small-scale prototype autos, machine learning, embedded systems, and automation. The purpose behind this undertaking is to show the design and development of a Cruise Control Car. This vehicle will be able to continue operating at the pace that is clearly visible on the road through speed limit signs detection, and even to notify a person who can intervene in case of an accident taking place. The whole project involves combining machine learning-based traffic sign recognition,

motor actuators controlled by an Arduino board, and accident detection that relies on ultrasonic sensors, in such a way making the road safety solution efficient and scalable for use in smart vehicles (fully and semi-autonomous) The system identifies standard speed limit boards (like 40 km/h, 60 km/h, 80 km/h, etc.) that the ML model is primarily a component trained to recognize. Using image processing and classification techniques, the system recognises these signs from the frames captured and via serial communication, it shares the identified speed limit with the hardware unit. The hardware section of the car is made of an Arduino Uno, motor driver, and BO (Battery-Operated) motors which give the capacity of very accurate speed control based on ML inputs. By turning the recognised speed limit into PWM signals for the motor driver, the car automatically keeps a safe and steady speed according to the recognised sign. Another Arduino Uno is assigned for accident detection, and it works in conjunction with the main system, with the ultrasonic sensors being the key elements of the whole process. It has been set up that the sensors will detect frontal or rear impacts or collisions, and they will be positioned accordingly. The moment an accident-like scenario is identified through violation of distance thresholds, the system triggers a buzzer. This module uses two ultrasonic sensors located in such a way to measure sudden frontal or rear accidents or collisions. If an impact is detected by crossing the distance threshold, the system turns on a buzzer for an immediate audible warning and transfers an alert via a Bluetooth module to a mobile phone. This procedure prevents delays in the relay of essential information, thus facilitating prompt action and enhancing situational awareness. The dual-circuit architecture, where one Arduino takes charge of speed regulation and the other one of accident detection, provides modularity, reliability, and ease of debugging. This sharing of duties not only increases the performance but also avoids the single microcontroller overload in processing. The combination of ML-based sign recognition and sensor-based accident detection offers a strong demonstration of intelligent vehicle control, safety automation, and IoT-based alert mechanisms. The project demonstrates how low-cost embedded platforms interfaced with machine learning can imitate the features of modern driver assistance systems, thus opening up access to such systems for engineering students, researchers, and robotics enthusiasts as a perfect educational prototype.

Key Words: Autonomous Vehicle, Traffic Sign Recognition, Machine Learning, Arduino Uno, Speed Regulation, Computer Vision, Accident Detection, Embedded Systems