The past few weeks have been a blessing for the Tesla community, as A refreshed Model Y performance was spotted in the Alps, and Tesla also started Robotaxi service in Austin, Texas, recently. Now, they announced their plans to change one of the key hardware components in their newer models of vehicles by changing their driver assistance systems and autonomous driving features.
The transition is accompanied by the introduction of the front-facing camera on the front bumper of the new vehicles as a part of a larger initiative in Tesla to enhance and develop its self-driving system based on vision.
This front bumper camera first appeared on the Cybertruck and has since been added to the revised Model Y, code-named Juniper. Although this hardware was seen months ago, Tesla had officially commented on its purpose and relevance just recently.
New Autonomous Vision
A recent illustration of the functions played by the hardware provided by Tesla also includes a combination of such functions as Autopilot and a new term the company has coined: Smart Summon. Recently, during an update of its vehicle specifications, Tesla showed its new, front-facing camera as one that will provide more visibility during parking or so-called Autopilot Smart Summon capabilities.
In January, when it further advanced Model Y, the company referred to the camera to boost the field of view on the automatic assisted driving and advanced Smart Summon. The new wording implies not only hardware advancement, but also a serious advance in the autonomous capabilities of Tesla.
The constant adherence to a vision-only system, which Tesla has used and termed as Tesla Vision, has been a feature of its autonomy strategy. In this system, ultrasonic sensors (USS) and radar are not considered or used, so the only power is the camera feed driving neural networks. The company thinks that this strategy is the one that best reproduces how human beings drive and see around them.
One Way to Get Rid of Sensors to have a PURE VIEWING Approach
When Tesla Vision came online several years ago, Tesla started to drop radar and ultrasonic sensors in its cars. Some industry analysts questioned this decision since most other automakers still use a combination of sensors, radar, and lidar in self-driving developments. Nevertheless, Tesla has been adamant that cameras are enough.
An important part of such a system is the occupancy network of Tesla. It gathers data provided by the vision sensor to build a 3D picture of its immediate surroundings, which may serve as a substitute for the ultrasonic sensor input.
Vision – Only the Belief of Elon Musk
As CEO, Elon Musk has always preached, vision is the key to autonomous driving. Regarding the new hardware and software updates, Musk underlined that cameras, combined with superior neural networks, can be even more efficient than human perception in many aspects.
Musk told attendees at the event that when your vision is active, it is better than the best human because it has eight cameras. It is as though you have eyes on the back of the head, by the side of the head, and three eyes with variable focuses staring in front. And receiving all this at superhuman velocities.
Musk refuses to accept that sensors could only provide temporary assistance in these vehicles, but eventually, they are crutches. Tesla wagered that the full-self-driving solution would feature a combination of software improvements and camera-only hardware, becoming dramatically safer than an average person.
