Apple's latest advancements on autonomous driving technology confirm that the company has focused on the software side of things. The technology giant is building autonomous systems that could be adapted to an array of different vehicles, with artificial intelligence (AI) at the center of it all.

The company's machine learning team recently published a new research pointing toward the same direction. In the research, the team introduced a new mapping system that could potentially be of use in future autonomous navigation, virtual reality, and housekeeping robots.

The new system is called VoxelNet, and its main goal is to improve on the data that is gathered by today's prominent self-driving component, the LIDAR sensors. In LIDAR, which stands for Light Detection and Ranging, the sensors bounce lasers off nearby objects in order to build a 3D image of the surroundings. The method offers more reliable depth information compared to regular cameras.

With VoxelNet, the 3D map is divided into different areas of interest by employing a number of discrete systems that can categorize them. In effect, the VoxelNet system can identify between cars, street signs, pedestrians, bicycles, and other objects.

Furthermore, the system compresses these steps into a single neural network to greatly improve its efficiency compared to today's technology. Based on benchmark tests performed by Apple's team, the VoxelNet outdoes its rival systems.

AI is expected to be at the center of the VoxelNet's ability to identify objects. After all, the company was very clear from the beginning that its foray into autonomous driving technology will be one big AI project. "We sort of see it as the mother of all AI projects," said Chief Executive Officer Tim Cook of the company's work on self-driving cars.

The company's focus on software systems also support the speculations that it would not manufacture its own vehicles, but rather create an autonomous driving system that can be installed on existing cars.