Tesla’s Autopilot Update Is One Step Closer to Full Autopilot

Tesla CEO Elon Musk said the company will introduce automatic identification of traffic lights and stop signs in the U.S. in the coming weeks,media reported. Tesla said in a feature presentation that cars can automatically stop at traffic lights and stop signs when automatic steering and traffic-aware cruise control is activated.

Tesla's Autopilot Update Is One Step Closer to Full Autopilot

Both features are included in Tesla’s base version of Autopilot, which is standard for all of the company’s vehicles.

Tesla’s traffic lights and stop sign identification have been in the works for a long time. In March 2019, Tesla released a new feature on the Auto Turn Parking Warning, which automatically alerts drivers when they approach an intersection and asks them to take over the control of an electric car.

The introduction of traffic lights and stop sign recognition and automatic parking will further enhance Autopilot’s safety.

Tesla said autopilot updates for customers in other countries will wait longer than U.S. car owners because of different traffic rules and related policies and regulations around the world.

Prior to that, more than 600,000 Teslas were equipped with fully autonomous-driving chips, which have up to 6 billion transistors that can perform 144 trillion calculations per second, process 2,300 frames per second of images at the same time, and have two such chips per vehicle that can process the same data at the same time.

It then applied for a patent on how to obtain training data from its vast fleet of customers to train its self-driving neural networks.

Tesla's Autopilot Update Is One Step Closer to Full Autopilot

Tesla's Autopilot Update Is One Step Closer to Full Autopilot

Tesla's Autopilot Update Is One Step Closer to Full Autopilot

Tesla's Autopilot Update Is One Step Closer to Full Autopilot

How to achieve autonomous driving

At present, the key technologies of autonomous driving are perception, planning and execution, involving sensors, data processing, machine learning, SLAM and sensor fusion, path planning and other fields.

Among them, the perception through the vehicle sensor hardware interaction and communication, planning is mainly responsible for the behavior of the car calculation, control is the electronic operation of automotive components.

Perception is mainly located through environmental perception.

In the environmental perception stage, it is necessary to obtain a large amount of information about the surrounding environment to ensure that the automatic car has a correct understanding of the body environment and corresponding decisions.

Environmental perception is the ability to understand the environment scene, for example, lane line signs and marking, traffic light recognition, traffic signal and sign age recognition, pedestrian vehicle detection, obstacle type and other data understanding analysis classification, positioning is the perception of the results of the post-processing, through positioning function to help the automatic car understand its location relative to the environment.

In the planning section, based on raw data and existing maps captured by the self-driving car sensor suite, the self-driving system needs to start planning the path from one point to another by building and updating specific environment maps through both positioning and mapping algorithms.

The latest development in machine learning is to efficiently process data generated by self-driving car sensors, reducing computing costs. In addition, advances in chip manufacturing and miniaturization are improving computing power that can be installed in self-driving cars.

5G high broadband, low latency will help network-based data processing to operate autonomously.

Execution is the system to control the vehicle according to the decision results.

The vehicle’s various control systems need to be connected to the decision-making system by bus, and can accurately control the acceleration, braking degree, steering range, light control and other driving actions according to the bus instructions issued by the decision-making system, in order to achieve the autonomous driving of the vehicle.

Perceived positioning is like the driver’s eyes, planning decisions are the same as the driver’s brain, and executive control is like the driver’s hands and feet. Executive control is the basis for autonomous driving to truly land.

Tesla’s fully autonomous driving plan

The biggest selling point for Tesla’s electric cars is its self-driving system(currently a level of secondary “assisted driving” level), and the company introduces upgrades to its self-driving systems at regular intervals.

The basic principle of the work of Tesla’s self-driving system is to accurately identify objects and objects on the road, and its driving computer will perform a variety of driving maneuvers based on these identified road conditions.

Tesla is scheduled to launch fully autonomous driving this year. Videos released by the company in the past show the vehicle being self-manipulated from the residence to the company without the driver mastering the steering wheel. This means that after fully autonomous driving, car owners can make phone calls, play with mobile phones and even take on business.

Musk even plans to set up an unmanned taxi company that would allow a large number of self-driving electric cars to drive freely on the streets and get passenger orders through mobile software.