Tesla recently released its first “Full Self-Driving” beta software update to a selected customer group, CEO Elon Musk tweeted. The CEO stated that more Tesla owners would get the software update soon as the week progresses, and there would be a “wide release” by the end of 2020. Only customers of Tesla’s Early Access Program will receive the software update, and the update will enable drivers to access Autopilot’s partially automated driver-assist system on the city streets. The early access program is currently used as a testing platform to help iron out software errors.

Musk stated that Tesla was looking forward and testing this software update “very cautiously” as the “world is a complicated and messy place.” In a letter to the investors, Tesla stated that its Autopilot team “has been developed on a fundamental architectural rewrite of the company’s neural networks and control algorithms. This rewrite will allow the remaining driving features to be released soon.”

Tesla Self-Driving Beta

The rewrite will also allow Tesla’s vehicles to understand their environment across four dimensions rather than two. This would also result in a dramatic improvement in the vehicle’s performance and faster software updates. Previously, Musk described a “feature complete” version of the full Self-Driving as enabling the vehicle to drive from home to the workplace without any intervention.

However, drivers will still require being ready to take complete control of the vehicle runs into a problem. Few tech experts have taken issue with the way Elon Musk revealed about the features in the past, stating that the CEO is muddying the waters by overselling and burdening a Tesla car’s capabilities.

Tesla, Tesla beta version, Tesla self driving car

Tesla Model Y

Tesla Self-Driving Car

The autopilot feature can center a Tesla vehicle in the right lane, even on the curves, and adjust its speed based on the vehicle ahead. The “Navigate on Autopilot” feature can suggest conducting lane changes to get around slower vehicles. It can also steer a Tesla towards highway interchanges and exits. Another advanced feature in the car can slow the vehicle at traffic lights and stop signs.

The company is yet to allow the customers total hands-off vehicle control at medium speeds, where drivers face more traffic signals, intersections, and other driving complexities.

The autopilot feature can’t perform tasks like, suppose the lane markers of a road are faded or missing, the feature can’t take turns. At that point, the driver must have a hand on the wheel at all times, or else the Autopilot feature will flash a series of warnings before ultimately disengaging off-track entirely. However, when all the features are in sync, it looks like the car is driving itself. But a driver would still be liable if the car makes a mistake or crashes.

Elon Musk also argued that Tesla’s self-driving advantage comes from having a large line-up fleet of vehicles on the road and functional- almost around 930,000. Those vehicles record situations and provide efficient training data to improve the neural networks required for developing AI software that powers self-driving cars.

Tesla’s approach to autonomous vehicles is focused mainly on computer vision or using cameras to recognize and understand the road ahead. “With around a million cars giving feedback and specified feedback on strange corner-case situations that usually cant be even created on simulations, that is a precious thing,” Musk stated.