Apple’s autonomous driving technology will integrate sensor data for perceptual decision-making

Apple has carried out research on autonomous vehicle system for a long time, including testing its developed vehicles in the United States. Although it is currently unclear what the company's ultimate goal is for this system, such as whether it will be used as one of the so-called "Apple car" features or licensed to other manufacturers, it is understood that Apple is still researching how to improve this technology.

On Tuesday, Apple obtained a new technological achievement called "sensor data sharing across sensor processing pipelines", in which Apple proposed that more collaboration can be carried out on the collected data in multiple processes within the autonomous vehicle system.

A typical sensor data processing pipeline, in simple terms, is to provide the data collected by sensors to a specialized processing system, which then judges the situation and relevant action plans, and sends them to other systems. Generally speaking, data from sensors will be limited to one pipeline and will not be affected by other systems, but Apple proposes to make fusion perception decisions based on the fusion of data from multiple sensors.

In addition, Apple has proposed that these data can be further refined, decisions can be made at different stages of each pipeline, and integrated together. These data can be shared between pipelines and can affect other processes, including raw sensor data, processed sensor data, and even data derived from sensor data.

By providing more data points, the control system will have more information available to use when creating action plans, which may lead to wiser decisions. Apple mentioned that these pipelines can be used to connect different types of data, for example, the first pipeline is used for image sensor data, while the second pipeline uses LiDAR data.

By combining these data, it is possible to make judgments that are usually impossible to achieve with just one set of data. For example, LiDAR data can determine distance and depth, but it cannot see color, which may be important when trying to identify objects on the road, and it can obtain color data from image sensor data processing pipelines.

Apple submits a large number of technical achievement applications every week, although these applications do not guarantee that Apple is developing specific features or products, they do show the field in which the company's research and development work is located.

Prev: Small Bluetooth has a great impact, and wireless connection in the 5G era cannot be without it!