VinFast is betting on autonomous driving through cameras and is already testing it

Autonomous driving presents many challenges and very demanding technology requirements. A large number of sensors, radars and systems are needed to capture information about the car’s surroundings so that the processor can interpret it and act accordingly. But some brands want to make autonomous driving more accessible and much cheaper.

VinFast is a brand from Vietnam that has proposed a simpler way to enable autonomous driving than using the more complex radar and sensor systems. The brand’s proposal relies on the use of cameras, which will form a complex system capable of fully mapping the environment and reacting according to the data collected.

The Robocar system

This platform is called Robocar and consists of seven cameras and a processor that performs the calculations based on the data obtained by them. This makes it cheaper than using sensors that are more expensive to implement, such as radars and LiDAR.

This equipment has been installed on two models, the VF8 and the VF9, and testing has begun in a controlled area of Hanoi. The start of these tests marks a before and after in the use of autonomous driving, since this system is much cheaper to implement and could be used in more affordable models than those that usually use radar and sensor systems.

From level 2 to autonomous driving

The test programme starts with an initial phase that will improve current SAE level 2 ADAS systems. From this starting point, the aim is to evolve toward fully autonomous systems. Of course, this system faces significant challenges, since without LiDAR and radar, algorithms must be developed that can interpret complex scenarios with many changes. It will rely only on visual information, which makes everything more complicated and demands a great deal of work from the processor.

A complicated environment

The chosen environment is important. In fact, Hanoi is a city with a great many motorcycles, which move in an apparently chaotic way and create many changes throughout each journey. This makes it much more complicated than driving in a Spanish city, for example.

The tests will be carried out over several months, so a wide range of data can be collected to focus the results on real traffic and to combine the use of cameras, their visual information and algorithms to achieve an advanced level of autonomous driving.