byADIT Free software for engineers and students
EngiFree
CARLA — The Complete Guide: An Open Simulator for Autonomous Driving, Sensors and Traffic Scenarios

CARLA — The Complete Guide: An Open Simulator for Autonomous Driving, Sensors and Traffic Scenarios

CARLA is an open-source (MIT) simulator for developing, training and validating autonomous driving systems. It lets you test perception, planning and control algorithms in a virtual world before moving to expensive and risky field trials.

What you get

Hardware requirements and who it suits

CARLA needs a powerful computer: a dedicated graphics card with at least 8 GB of memory and about 20 GB of free disk space. It suits researchers, graduate students and automotive and transport engineers developing driver-assistance systems. For city-scale traffic flow simulation it is not the right tool.

Related tools

For microscopic traffic simulation of an urban network — SUMO. For the robot and vehicle software itself — ROS. For general robot simulation — Gazebo.

The bottom line

For testing autonomous vehicle algorithms before they reach the road, CARLA is the leading open simulator — provided you have suitable hardware.

Further reading

מדריכי AI למהנדסים ולמשרד, ב-5 שפות: adit-ai.com ←