Smart cars have one Achilles Heel that seems so trivial. Parking lots. Like human drivers who cannot pillow park or back up into a slot easily. In parking lots, they are prone to error. This prompted Tesla to create "Smart Summons" like those Sci-Fi flicks in the movies. Since the over the air (OTA) update that has been used 1 million occasions. This data will be applied to machine learning that will enable Tesla cars to tackle a parking lot better.
Automated cars are perfect for highways that go on a straight line. Enter the parking lot. If smart cars had emotions, this would be parking blues or anxiety attacks. Experts have remarked that when and if any Tesla or similar smart car manages any parking lot. No bumping, collisions or dings while getting out of a slot. It will be worth getting any smart car that has smart summons.
Another thing about smart cars is that highways are wide open. Machine learning needs only to know, when to engage the auto brake, adjust the car speed in adaptive cruise control while driving. Fewer obstructions are in front of the car, so easy peasy to adjust around them. Maybe a few pedestrians or several bikers at most.
Start throwing them in a crowded parking lot where everything is chaos. Car sensors pick up and detect many obstacles that are akin to information overload. When machines learn that will cost the owner. Tesla smart cars will be scolded for peeing on the rug. Functions like smart summons, although the hardware needs to get more brains.
One professor at UC Berkeley remarked that "localization" is difficult in a parking lot, whether it is open or underground. Cameras are not ideal, without precision. Visual detection cannot be the same in all conditions, it has limits. Equipping a laser sensor or lidar that is used on some test cars like Waymo is expensive, guarantees better electronic sensing. To guide autonomous cars in parking lots, they need better sensors.
Underscoring the need to use any collected data is Tesla's CEO stating that it is relevant, and still in the testing stage. Data gathered from the car's sensors is to assess safety performance which drew comment from agencies. Information gleaned from the use is the most important component to improving how smart cars work.
Those using smart summons have observed that Tesla could not differentiate between asphalt or grass surfaces. Almost hitting another car in the lot. For the reputation of Tesla cars, this was a bit off. However, as explained that it is not perfected yet. These circumstances gave valuable data to make cars smarter.
For sure Tesla's Smart Summons OTA update will be fixed, and bug-free eventually. Mastering the parking lot is the key to operate smart cars flawlessly in any situation. Is a Tesla that parks and get summoned without hitting anything a good deal? Or stick with parking yourself.