Apple’s Tesla Model X Rival ‘Project Titan’ Car To Feature ‘Confidence’ Algorithm

A new algorithm just might help self-driving vehicles in Apple’s “Project Titan” recognize roads and road objects faster, allowing them to respond more accurately and safely.

Self-driving vehicles are heavily dependent on the data their sensors collect on the road. By quickly processing the data collected, these vehicles are able to make safer decisions on the road faster.

Faster processing of road data, however, requires much, Apple Insider noted. Companies who make self-driving cars can equip their autonomous vehicles with more powerful sensing and processing technology to make faster computations, but this will come with increased power consumption and higher costs.

Alternatively, these companies can design the self-driving car’s system to collect data selectively, so that it won’t have to process as much data as the one mentioned above. This can be done with less equipment, and will not consume that much power as well.

The latter is what best describe’s Apple’s plans for self-driving vehicles in their secretive “Project Titan.” In a patent recently filed at the USPTO, Apple described a new technology that combines depth data collected by active sensor devices with image data collected by passive sensor devices.

Data collection

According to the patent titled “Depth perception sensor data processing,” the sensor data processing system collects data using both passive and active sensor devices, and uses the data to generate an overall map of the environment immediately around the vehicle. This map becomes the initial model.

The system then collects another set of data using more powerful equipment like LiDAR (light detection and ranging) sensors and compares this to the initial model to create a new model. The system will keep adjusting its model to accurately represent roads and road objects near the vehicle.

Confidence

The amount of adjustments made to the model is called “confidence.” Once the system establishes enough “confidence” that the model is accurate, the model will then be added to the self-driving system.

Apple’s “Project Titan” is expected to benefit much from the “confidence” algorithm, which uses less hardware and resources to quickly generate models accurately representing the road -- that ultimately leads to faster responses and safer rides.