AP is now so good 99% of the time that I notice when it isn't performing at its best. My question is - is this last weak spot fixable with the current hardware? I'm wondering if it's a limitation of the imaging chip used in the camera. Or, if as the high precision GPS maps get gradually "built" by the fleet, if the low contrast, high glare situations will get fixed?
These situations are noticeably better after only three months of the software being in service - but the car still seems to wander a bit on certain stretches of I-10 in Southern California where a series of factors stacks up simultaneously, making for very challenging conditions:
1 - Low sun ahead, causing glare
2 - Poor lane markings
3 - "Shiny" pavement, for lack of a better term - 'causing even more glare
4 - No cars to my sides or ahead for the car to use to get extra information on its position.
In this situation I've noticed the car wandering - not nearly as bad as 7.0 but still some "hunting."
Since it is blind to the rear past 16', a situation has arisen a couple times where I've felt the need to grab the wheel because a car was approaching at high speed in an adjacent lane from behind and I didn't want the Tesla to swerve toward that car's lane at the wrong moment.
In any case - what is your speculation on whether this behavior can/will be further improved on the current hardware suite?
These situations are noticeably better after only three months of the software being in service - but the car still seems to wander a bit on certain stretches of I-10 in Southern California where a series of factors stacks up simultaneously, making for very challenging conditions:
1 - Low sun ahead, causing glare
2 - Poor lane markings
3 - "Shiny" pavement, for lack of a better term - 'causing even more glare
4 - No cars to my sides or ahead for the car to use to get extra information on its position.
In this situation I've noticed the car wandering - not nearly as bad as 7.0 but still some "hunting."
Since it is blind to the rear past 16', a situation has arisen a couple times where I've felt the need to grab the wheel because a car was approaching at high speed in an adjacent lane from behind and I didn't want the Tesla to swerve toward that car's lane at the wrong moment.
In any case - what is your speculation on whether this behavior can/will be further improved on the current hardware suite?