4 ms·
That's basically true, but there's no reason to limit it to Autopilot. Although the high points of AP are better than any other system that I've tried so far. I
by jsight 5y ago
That's basically true, but there's no reason to limit it to Autopilot. Although the high points of AP are better than any other system that I've tried so far. I've seen it do amazingly well in the rain at night, and in the rain on a road with faded markings that were hard to pick out. I've also seen it thoroughly confused by double-dashed lines at hov lanes.
I'll still take that over the ones that beep proudly to tell you that they are about to fail to maintain their lane (hi ProPilot).
- notJim 5y ago> I've seen it do amazingly well in the rain at night Not anymore, since they removed the radar. Rain seems to really interfere with the vision based system, and apparently auto high beams are required at night, and they flash constantly.
- bordercases 5y agoWhy did they remove radar?
- brianwawok 5y agoLong term: Vision is better, and less prone to false errors (see all the talk of phantom breaking). I think it was on the way out by maybe 2022 Short term: Global chip shortage made supply tight, and it was either remove it and rush out with the no radar code branch, or stop shipping cars. Long term I think it's a fine decision, but short term it's kind of a big thing to rush.
- ketamine__ 5y agoDoes it give you pause that practically every other company pursing self-driving cars disagrees with Tesla?
- brianwawok 5y agoSure, in 5 years we will see what shakes out. I am not taking my entire savings and investing in tesla stock or anything. That said, i want a cool fast BEV that makes highway driving easier. Right now, Tesla is the only option. A few others are close. If someone else beats Tesla to FSD with another system, I will be happy to go that way on my next car. Working FSD is worth a lot to me.
- ffhhj 5y agoTo buy more bitcoins.
- zeusk 5y agolove react because HN only does upvotes. I feel like they're straying from their mission. They did really well with moving the market but now that the market is moving, I have a hard time believing Tesla will stay at the top.
- benhurmarcel 5y agoCost saving
- simondotau 5y agoIt's worth noting that Subaru vehicles have been doing vision-only adaptive cruise / lane keeping / emergency braking for many years, which they brand Eyesight. In my personal experience its adaptive cruise is as good or better than most other cars I've driven which use radar. (Admittedly the reasons why the Subaru impressed me was more for its programming rather than absolute precision. I liked how smoothly it transitioned between following and open road ahead. Much less jerky.)
- interestica 5y agoRain really interferes with human vision based systems as well. It's insanity that we don't better design road markings to better compensate. It's often the case where old painted-over lines are more visible on wet roads than the current markings. Making road information systems standardized and better for humans works out great for autonomous cars.
- jsight 5y agoYeah, I know. The vision based one seems to be more limited in the rain at the moment. I'm questionable as to whether this will be a long term problem, since it does so well at picking up lane line markers. It seems like a solvable issue. The high beams are a different story. I only like to use auto-high-beam on rural roads as they tend to be problematic on busy highways in Tesla and my non-Tesla. But requiring them even on highways would be a big problem that I don't foresee them solving soon. I'm worried that this is going to take a long time to fix. IMO, they actually work pretty well on rural roads.