A Systematic Mapping of Autonomous Vehicle Prototypes: Trends and Opportunities

Autonomous driving vehicles (ADVs) are the next significant evolution in transportation systems. Using machine learning techniques supported by strong processing units and sensors, the driver would be removed and the vehicle would drive itself, bringing countless advantages to the commuting system. There are too many players developing their own ADV prototypes all over the world. For this reason, this paper aims to use a systematic mapping study (SMS) to identify trends and patterns in ADV development. From a set of academic databases, a total of 68 papers were selected from a total of 548 after applying inclusion and exclusion criteria and snowballing techniques. The results showed the technical characteristics of the platform required to develop real-size and mini ADV prototypes. In addition, the paper explains the detailed software and hardware architectures and their main aspects and requirements. Finally, the paper discusses the necessary set of sensors which are required and their characteristics and main suppliers. It was possible to map the publications distribution based on venues, year, authors' affiliation, and most prolific countries. This work can guide researchers to design their ADV prototype or identify development bottlenecks that could lead to narrow research and innovation possibilities in autonomous driving vehicle development.