Tesla has patented a system for improving its full self-driving (FSD). The described invention detects dynamic objects using vision-based systems and creates their path prediction in order to make autonomous driving as safe as possible.
FSD is Tesla's biggest goal. For several years, the manufacturer has been developing Autopilot, which will be able to fully drive the car. At the moment, the company is working on FSD Beta, which can already drive a car with minimal driver interventions. However, it requires constant supervision from the human driver. The process of creating software that will make cars fully autonomous is extremely difficult, but Tesla is making steady progress in this.
On February 23, 2023, the patent, “Detected object path prediction for vision-based systems,” was published, which discloses the utilization of a set of inputs from vision systems to generate simulations or predicted paths of travel for dynamic objects detected from the vision systems.
The company explains that traditionally, vehicles have physical sensors that can be used to provide input to control components. For many navigational, location, and safety system, the physical sensors include detection-based systems, such as radar systems, LIDAR systems, etc., which are capable of detecting objects and characterizing the attributes of detected objects. However, the use of such detection systems can increase the cost of manufacturing and maintaining vehicles. In addition, for example, during rain, fog, and snow, such detection systems may not be suitable for the detection or increase the number of detection errors.
Therefore, in its cars, Tesla relies primarily on cameras that are relatively inexpensive and “see” better in various weather conditions and environmental features. To address some of the shortcomings associated with the use of sensors, aspects of the published patent correspond to “utilization of a set of inputs from vision systems to generate simulations or predicted paths of travel for dynamic objects detected from the vision systems. That is, the service can "process the set of inputs (e.g., the associated ground truth label data) collected from one or more vision systems (or additional services) to identify predicted paths of travel for any dynamic objects detected from the captured vision system information Typically, a plurality of predicted paths of travels can be generated such that more than one path of travel may be considered to meet or exceed a minimal threshold.”
In fact, vehicles with machine learning algorithms and without any additional detection systems such as radar detection systems, LIDAR detection systems, etc., using the systems described in the patent, can provide detection of objects and predict their path of movement.
© 2023, Eva Fox | Tesmanian. All rights reserved.
We appreciate your readership! Please share your thoughts in the comment section below.
About the Author
Eva Fox joined Tesmanian in 2019 to cover breaking news as an automotive journalist. The main topics that she covers are clean energy and electric vehicles. As a journalist, Eva is specialized in Tesla and topics related to the work and development of the company.