In order to overcome the shortage of transportation, modern technologies are proposed by the researchers. In particular, there is a phenomenal burst of interest in smart vehicles. Despite the benefits of smart vehicles, significant challenges remain to be addressed. An important problem is sign recognition in the presence of noise (sever weather condition). Moreover, redundancy in data recording ensures a more reliable detection. Besides using cameras, LiDARs are also used in autonomous vehicles. In this thesis, we propose solutions to address recognition as well as measures to improve recording redundancy. This main three stages are: 1-implementing a testbed which is a Toyota equipped with necessary software-hardware, 2-proposing a neural network algorithm to recognize the road and traffic signs even in the presence of 90% noise (sever weather condition), 3- equipping the vehicle with LiDAR system that improves data recording capability. All the stages are supported by practical and simulation results.