Xpeng VLA 2.0 Test Drive: Evening Rush Hour Guangzhou China - Xpeng GX

[00:00:00]Hey everyone, welcome to our second video of VALA 2 in the Exponent GX. We are already activated on VALA. It's stopped at the boom gate until the we get clearance once we um pay the payment >> for parking.

[00:00:16]>> So, it's smart enough to know there's a boom gate.

[00:00:18]>> Yeah, you can see it on the screen there.

[00:00:19]>> Y >> feel free to come closer and have a look what's happening on the screen.

[00:00:28]And I think these are all paid with QR codes, isn't it? Scan your phone.

[00:00:33]>> Well, use your phone to scan the QR code.

[00:00:36]>> Do I need to nudge it forward?

[00:00:39]>> Should I push the >> A little guess?

[00:00:40]>> Little guess. Y >> just a bit. Yeah, just to nudge it just like Tesla. Sometimes it gets stuck.

[00:00:46]>> Just Well, I think this is probably just more confirmation.

[00:00:49]>> Yeah. But so we are actually approaching peak hour traffic. It is 4:00 >> and we're doing an airport run.

[00:00:57]>> This person's walked right across, but >> y >> all good. The car was reacting appropriately.

[00:01:02]>> We are getting out of this lot. Um, we've just cut in front of a bus.

[00:01:07]No problem.

[00:01:12]And now it's scanning the road and showing you these three lanes are all good. Don't go in there. That's good.

[00:01:19]It's very raises confidence in me knowing the car knows what to do.

[00:01:25]touching.

[00:01:26]>> So yeah, in just to remind you in China that we have to um hold on to the wheel even though the car doesn't need it, but just for regulation, the cameras do pick up the fact that we may not be touching the wheel.

[00:01:36]This is my position just holding the wheel like that.

[00:01:53]And now it's telling us the car is planning to go to the right lane eventually cuz that's the land it wants to be in.

[00:02:02]I love this map. this heating up.

[00:02:17]I >> feel like um I think a a lot of driving in China is is different obviously to Australia.

[00:02:22]>> It's full of edge cases.

[00:02:24]>> There's really good vis visualization of these um little >> pillars here as well. The safety pillars. Yeah.

[00:02:31]>> Multiple. Yeah.

[00:02:32]>> Yeah.

[00:02:35]I want to call them witches cones, but they're not um they're not the right shape.

[00:02:39]>> There's a truck entering.

[00:02:41]It's sort of going around it.

[00:02:43]>> Yeah. Yeah.

[00:02:46]>> Great. High quality high definitionition cameras.

[00:02:55]I think seeing the car plan ahead is is really good for for the human monitoring it.

[00:03:01]>> Yep.

[00:03:04]Certainly something um I think Tesla FSD could learn from just to give like you know you can say all you want the software knows what it's doing but you're still monitoring as a human so you want that confidence as well.

[00:03:21]>> It helps people to feel relieved isn't it? Yeah.

[00:03:24]>> And that's what it's about. It's actually convincing the human population that the software is is good enough and that it shouldn't be understated.

[00:03:32]>> No. And visual feedback is always important and redundancy in visual feedback is important as well. So just because there's one, you know, one method of showing something doesn't mean you can't have multiple methods of showing the same thing at the same time.

[00:03:45]>> I mean, a lot of early adopters now with FSD, you know, it's obvious to them. And I get that, you know, you've researched the software and the history behind it, what it's doing. You get it. I know you get it. But we still need to, you know, convince the populace that it's safe.

[00:03:59]>> I think there's value in doing that.

[00:04:17]>> See, now I can even see the barriers coming up ahead on that overpass.

[00:04:21]>> Yeah, >> it's fantastic.

[00:04:24]knowing they're like solid barriers there and telling the human that the computer knows what's going on.

[00:04:31]>> There's another tunnel coming up, I think. Looks like >> and again knowing the elevation change.

[00:04:40]So confidence we're going into the tunnel.

[00:04:42]>> Yeah, >> quite a bit emerging again into these two lanes.

[00:04:49]>> Can Australian roads be this well mapped? That's the question, isn't it?

[00:04:54]>> Look at that.

[00:04:55]>> I just looked behind me because I just to make sure that car wasn't it was of course >> and I was looking to the side, but >> and again I'm a foreigner here, so I don't know these roads at all. I'm complete tourist yet. I'm feeling quite at ease.

[00:05:13]>> You want to tap out your heart rate like what I did?

[00:05:30]to be fair. I just had a nice tea.

[00:05:32]>> Yeah. Yeah.

[00:05:33]>> I'm a bit caffeinated.

[00:05:34]>> I reckon it's 70s.

[00:05:43]>> And I think there's merit in continuing to stress test any self-driving system in the city.

[00:05:49]um especially in busy cities like Guangjo, like Sydney, >> London, anywhere around the world.

[00:05:58]>> And you know, we were lucky enough to be able to test the previous version of this, which was XNGP last year in >> it was just in November, I believe, wasn't it? Was it? And um >> it was already impressive then, but there were a couple of times, I think, that we intervened, isn't it? Um, and that was in a much shorter space of time. This one is like generationally ahead.

[00:06:23]>> Yeah, it's leaps and bounds ahead.

[00:06:26]>> Feel a lot more confident.

[00:06:27]>> It's incredible how much the tech has changed in 7 months, isn't it?

[00:06:34]>> The visualizations are just so good.

[00:06:37]>> Yeah.

[00:06:38]So, it does need the brain power to support it, doesn't it, in the car itself.

[00:06:45]>> Mhm. So you will need a newer version.

[00:06:48]>> Yeah. Was it three three touring chips or something?

[00:06:55]>> Nice. Just see the truck behind me.

[00:06:58]>> These are slowly really slowly cars.

[00:07:04]>> Picks the pathway.

[00:07:08]>> He needs to go back somehow that green lane eventually.

[00:07:22]One thing that may not come across on camera is just how incredibly smooth the acceleration and braking is in this kind of heavier traffic, isn't it?

[00:07:36]>> Yeah, it's not nauseating at all.

[00:07:37]>> Yeah.

[00:07:40]>> So smooth.

[00:07:42]smoother than any human could do for sure.

[00:07:44]>> Definitely. Especially in this slow traffic.

[00:07:46]>> Yeah.

[00:07:47]>> Look at the gradation of the green into yellow coming into a slow zone for this part, but then green again. So that's really very discreet how it's mapped out like that again.

[00:07:57]>> I I would not want to be navigating this intersection.

[00:08:00]>> We have to actually go into that bit. So >> oops. That was >> we are close to the concrete but no we're good.

[00:08:14]>> Wow.

[00:08:16]>> Oops.

[00:08:20]>> Okay. I pushed the brake then.

[00:08:21]>> You pushed.

[00:08:22]>> Yeah, I did. Yeah. Was that the right thing to do to intervene?

[00:08:25]>> To start it again.

[00:08:28]>> Should I have let it go?

[00:08:30]>> The truck was coming.

[00:08:31]>> Intervention one. Yeah, that seemed >> that that's another one of those situations where I would rather not trust.

[00:08:37]>> Do you feeling he's going to break or not?

[00:08:40]>> Not really. No, maybe.

[00:08:42]>> You feeling?

[00:08:43]>> Me? Well, >> do you feeling >> I wasn't confident enough. So, >> it's And again, see, that's a cultural thing because like back home and because also we went to the vehicle management office and had to watch all those videos of the car of the trucks just like um taking out all the other vehicles on the road. So, um, as a passenger, I much preferred that, um, that that Tom just intervened in that situation. Yeah. I'd rather not find out the hard way. Yeah.

[00:09:10]>> I mean, that car that truck was very close.

[00:09:12]>> Very close.

[00:09:13]>> Yeah.

[00:09:14]>> It's not my car, so I didn't want to risk it.

[00:09:15]>> Yes.

[00:09:18]>> Yep.

[00:09:19]>> But I feel he's being a little bit slowly slowly, but he's fighting with >> Yeah. And that's that's the thing, isn't it? Like, we don't know >> the personality of the of that truck driver. It probably would have, but >> why take the risk?

[00:09:33]>> Maybe at least 15 cm until 5 cm here.

[00:09:38]>> Yeah, >> I'm not playing chicken with a truck.

[00:09:41]>> Would you would you have stopped as well or would you let it go?

[00:09:43]>> I never dropped like this. So close.

[00:09:48]>> Okay, good. Sorry.

[00:09:50]>> Never met close like this.

[00:09:52]>> Okay, cuz you're a local. I I have to ask you what you would have done.

[00:10:00]I mean, you know, the tolerance is a lot uh greater here in Australia. I would have break much sooner.

[00:10:05]>> Yes. Yeah.

[00:10:10]>> Sorry, that was quite a sharp break.

[00:10:15]>> But if I drive by myself, I'll stay far away from this type of car.

[00:10:20]>> Of course. Yes.

[00:10:31]And that's the other thing like it's one thing I think for for people who are watching this video to go ah I wanted to see what would happen in that situation when you are in the vehicle in the car it is a completely different story >> slow.

[00:10:52]So, it's about 2.6 km worth of this. Is it 9 minutes worth?

[00:10:57]>> 9 minutes.

[00:11:00]>> Do we want to see us in slow moving traffic for 9 minutes?

[00:11:03]>> I don't know.

[00:11:05]>> Should I pause it?

[00:11:06]>> Maybe pause it there.

[00:11:07]>> Yeah. All right. We will we will be back um when we're no longer in this standstill traffic.

[00:11:16]>> Well, it's not stand still. Slow moving traffic. Have >> a break. Yeah. See you very shortly.

[00:11:22]Okay, we've just saved you about 10 minutes worth of slow moving traffic.

[00:11:25]Nothing really happened. I mean, cars were merging in and out, which is fine.

[00:11:28]We handled it pretty well. So, you can watch us struggle in the slow traffic for the last minute.

[00:11:34]But I'm just thinking about that incident again with the hard braking. I felt like I didn't push it that hard, but it almost felt like the car maybe simultaneously emergency broke with me.

[00:11:44]>> Don't know. I guess we won't know.

[00:11:48]But see how it's um the gradation is now turning yellow pretty soon. We know it's going to speed up again.

[00:11:53]>> Mhm.

[00:11:54]>> Can see what happens with all the merging cars.

[00:11:55]>> I was going to say I mean it can it's got it's got them all on the visualization so it's aware of them.

[00:12:00]It's already doing its calculations.

[00:12:05]>> So we should start moving slowly soon.

[00:12:11]>> Yeah. I can see up ahead the car's moving ahead.

[00:12:13]>> Yeah. See who does it let through and who does it not let through.

[00:12:17]>> Yeah. This is nice having this. um how much left in traffic. I'm not sure that the Tesla tells you that. I think we rely on ways back home to tell us how much longer in slow traffic, >> which is why we've normally got our phone mounted in front of the um >> Yeah.

[00:12:33]>> steering wheel. Y >> So not in front. Do you consider that I mean it's in front for the car, but it's behind from your point of view, isn't it? So whichever way you think of it.

[00:12:46]>> Okay, now we're moving again. H.

[00:12:51]>> There's green coming up too.

[00:12:52]>> You can see how long left before we reach 60.

[00:12:56]>> There's an miss.

[00:12:58]>> There you go. Look. Let it in.

[00:13:01]>> Yep.

[00:13:01]>> Yep.

[00:13:03]>> Okay. And then green coming up.

[00:13:05]>> Yep.

[00:13:08]>> So, this has got ways kind of built into the system already, right?

[00:13:13]I wonder if it reports traffic hazards up ahead.

[00:13:17]Look, this ion is kind of like in our lane, but you can see that on the vis visualization as well.

[00:13:24]>> Yeah.

[00:13:25]>> And the car is taking that into account.

[00:13:27]>> Yeah.

[00:13:30]>> Okay.

[00:13:44]I guess we're we're sort of reaching the point where if it's got enough data and enough computing power, it should reach a point where it's actually going to be better at making decisions than humans, isn't it? In in most situations.

[00:14:05]I mean apparently you can see quite far ahead whereas I wouldn't know where the turnoff is right >> again as a tourist >> I am itching to test this in Australia >> to be able to test this in Australia would be insane absolutely incredible >> we are waiting with beta breath good to have a a competitive FSD so we can compare >> yeah cuz I believe there's some early testing in Europe is it with >> Yes.

[00:14:33]>> Yep. but not in any other jurisdictions so far. So, it's good to know that like, you know, it's Xfang's not just in China for this, that they are actually being able to test in overseas markets.

[00:14:48]>> Oh, yeah. They're the first self-driving to leave China globally.

[00:14:53]>> They are, aren't they? Are they?

[00:14:55]>> Yep.

[00:14:56]>> Well, yeah, I think so.

[00:15:00]Give >> FSD a run for its money. I'm all for it.

[00:15:07]Can I say this is night and day from our um previous airport transfer in Guanto which was in a a DD.

[00:15:19]Um and it was in a car that was a very very old car which back home would have been a zero star and cap rating car. Um, and it was the driver was doing 130 kilometers an hour.

[00:15:34]>> It was crazy.

[00:15:35]>> And yeah, >> this is not more.

[00:15:36]>> It was um extremely scary.

[00:15:38]>> It was very impatient driver.

[00:15:39]>> It was extremely scary. This is not scary at all. This is lovely. This is You know what? I'm going to say it. This is simply lovely.

[00:15:51]It's going a little bit slow, but it is 50 kph where it's, you know, >> it's obeying it's obeying the rules even if the other cars aren't.

[00:16:01]>> You you can um you can set it to do over.

[00:16:06]>> Yeah, we are now 10k off offset, >> which it is 61.

[00:16:16]>> So, we're back into slow traffic again.

[00:16:19]slowish.

[00:16:21]Okay, it's quite interesting. We're merging like both sides of the road being undercut. We need to get back in that that side. So, okay, that wasn't too bad.

[00:16:33]Okay. And then it's it'll speed up again.

[00:16:38]Yellow bit.

[00:16:42]>> Yeah. And it knows we can go up to 100.

[00:16:46]thought that we would be able to do so >> today. Um, and it also knows that well you it's 100 but we can go up to 10 over because that's what we've set it for.

[00:16:57]Um, and you can set it as either a percent or as kilometers per hour.

[00:17:09]I feel like if that was an option back home, I would choose percent.

[00:17:14]>> Because going, you know, 44 in a 40 zone, 55 in a 50 zone.

[00:17:20]>> I think it's the options were 5% 5% or 10%. So >> I just want to show that it actually knows that that's an ambulance as opposed to >> So it can identify emergency vehicles.

[00:17:32]>> Mhm.

[00:17:33]>> Yeah. It's differentiated from everything else. That's cool. I I mean, we haven't had an opportunity, I don't think, to test that yet.

[00:17:40]>> 14. Yeah, I've not seen that before.

[00:17:42]>> Yeah, I'm sure it's probably capable of doing that, too. But >> that's just good to show.

[00:17:50]>> Okay, we're coming up to the green bit, so we should be moving quickly again.

[00:17:54]>> One minute left in the yellow bit.

[00:18:01]Boop. Boop.

[00:18:17]can see a hint of green there just around the bend which I would not be able to see >> at Especially if you're commuting like this every day, right? It's actually quite useful.

[00:18:49]down to like stop start stop.

[00:18:56]>> There we go. Onto the green.

[00:18:57]>> Yes.

[00:19:14]Will we ever get to a stage in our lifetime where we can fall asleep when we are on the road?

[00:19:19]>> Yeah.

[00:19:22]>> Never say never, I suppose.

[00:19:25]Well, we know that at least there are safety features in the Xpang um for if a driver does become incapacitated or falls asleep, which we got to see um like a little video clip of.

[00:19:44]Um basically what it does is that it tries to alert the driver to wake up.

[00:19:50]Um, and if that doesn't work, it'll slowly and safely merge over um to the shoulder um pop the hazards on and call emergency services basically, which is um yeah, pretty awesome feature because yeah, someone might have a heart attack or something like that. Um and that's certainly um yeah, really good kind of like software programming. I would not be testing that today.

[00:20:21]>> Nope.

[00:20:25]>> That video might be on YouTube.

[00:20:27]Actually, it was tested by a couple of um I believe they might be Russian content creators. Is it?

[00:20:33]>> So, >> now it's coming up to speed.

[00:20:36]>> Now we're up to speed. Look at this.

[00:20:38]Look at this.

[00:20:38]>> Going around that car.

[00:20:39]>> Interesting. It kind of Yeah. went around.

[00:20:41]>> How fast am I going? Oh, here. Here we go. Instrument cluster.

[00:20:47]So used to FSD now. I have to look for the speed here. But no, it's right in front of me.

[00:20:52]Let's see what it does. It's quicker than me. I just going to say what's how's it going to deal with a slow truck.

[00:21:00]Immediate overtake.

[00:21:04]You can see the toll booth beneath beneath beneath us >> from here too.

[00:21:13]I like how there's like a little like safety heat map as well when you pass the truck. Um, did you see how it went a little bit yellow?

[00:21:22]cuz that's honestly how I that's what my anxiety level does when I pass a truck.

[00:21:47]>> Yeah. I mean, we don't I don't do that much highway driving in Sydney. Um, but I can see the value of FSD like this or VA2.

[00:21:56]>> If you have to take a motorway home or go on the freeway for a bit to drive home, be very useful.

[00:22:26]Yeah, no dramas with that taxi just um merging over.

[00:22:38]>> I love how we can also read who's got their brake lights on. Have you noticed that?

[00:22:44]>> Oh, from the other cars. M.

[00:22:46]>> Oh, yeah.

[00:22:46]>> Keep an eye out for it. Notice that.

[00:22:49]>> Point up.

[00:22:50]>> Like, you wouldn't think that on this um on this road people are breaking, but we might see it a little bit up ahead.

[00:23:00]>> It's going pretty quick now.

[00:23:02]>> Yeah.

[00:23:02]>> 113.

[00:23:06]We were going to have to break. So, you can see our brake lights came on.

[00:23:39]Oh yeah. Did you see that? The um that white car.

[00:23:42]>> Can see it on this one.

[00:23:43]>> Yep. Yep.

[00:24:02]What was it saying?

[00:24:04]>> Uh 2 kilometers later we'll go to resting area.

[00:24:11]>> Okay.

[00:24:12]>> Like if you need it is it >> is a rest >> rest stop. Okay.

[00:24:21]Neither of those two cars, consecutive cars, indicated.

[00:24:26]Um, but the vis visualization showed that it knew exactly what it was doing.

[00:24:32]>> Yeah.

[00:24:34]>> Yeah. It's already clocked the car's way ahead.

[00:24:36]>> M. Yep.

[00:24:48]I haven't seen that black vehicle indicate once.

[00:24:55]Still hasn't indicated at all.

[00:24:58]>> Not the one in front of us. The van.

[00:25:00]>> Oh, the van.

[00:25:01]>> H >> van.

[00:25:02]>> Yeah, that van. Yep.

[00:25:05]Can >> spot the broken line and unbroken line.

[00:25:10]>> See what it does. Does it get around this jersey?

[00:25:35]Okay, just getting around the slow car.

[00:25:40]Wonder if I can move this here. Yeah, I don't want to block.

[00:25:43]>> I love the way the blind spot monitor is really more for >> That's better.

[00:25:48]>> Yeah, good.

[00:25:49]>> It's more for our own anxiety relieving, isn't it? Because the obviously the car can see a lot better.

[00:25:55]>> Yeah. Oh, this is good. It's like Tesla's FSD moving up over there.

[00:26:02]>> So, you can see the other side of the road.

[00:26:06]>> Yeah, there's service road here too.

[00:26:12]>> Even the trees.

[00:26:14]>> I think the trees aren't accurate. I got to say well >> the necessary animation. That That would be a waste of computing power.

[00:26:22]>> Yeah.

[00:26:26]>> Just spot some of the buildings. I wonder if that's map data.

[00:26:29]>> That's building.

[00:26:30]>> Yeah.

[00:26:30]>> Yeah. I would I would assume that's nap data.

[00:26:32]>> Yeah. I see the heat map you're talking about.

[00:26:35]>> I know, right?

[00:26:38]>> Let's see another example of it when we pass.

[00:26:42]Oh, no. We're not going to >> going to pass.

[00:26:46]>> No. Not safe. It's not going to >> No, >> it's not going to merge.

[00:26:50]>> No, >> it's good.

[00:26:54]>> The calculations are being done.

[00:26:55]>> We might see that little yellow thing when we pass. See?

[00:26:58]>> Yeah.

[00:26:58]>> Look at that >> 123.

[00:27:11]>> So, is the speed limit really 120? M I saw it as well on the bill on the um street signs there. 120 >> cuz I choose 10 kilometers >> up.

[00:27:31]>> Yep.

[00:27:38]>> So if we were on a 10% offset, it would be happily going up to 133 km an hour.

[00:27:44]>> Mhm.

[00:27:56]It's rushing up to this car.

[00:27:58]>> Yeah. But it didn't like it didn't sort of get up close and then suddenly break.

[00:28:01]It was really smooth.

[00:28:05]>> Did it feel lurchy to you at night? Not at all.

[00:28:28]You can see the speed limit up on that board there.

[00:28:32]>> 120 is correct.

[00:28:33]>> I'm going to quickly pop over there.

[00:28:49]Show the average speed.

[00:28:52]>> Oh, so far on this drive.

[00:28:54]>> Yes, this is average speed. We cross the highway because this highway maximum speed is 120. So this is average speed is 94. We drive. Our driving is average.

[00:29:09]Okay.

[00:29:11]>> Okay.

[00:29:12]It's good because some high speed in China uh cameras don't work. But in this area from A to B, you need to drive slowly than 120. So the average Okay.

[00:29:38]>> So, should we Oh, actually, we're going under that now.

[00:29:42]>> So, he show you the average speed.

[00:29:47]>> So, does the does the camera pick up average or >> average from the start to the end?

[00:29:52]>> They're all average speed cameras, right? Rather than at that point. Okay.

[00:29:58]>> Interesting.

[00:30:00]So that's where it's not good to have an offset ofund the well a 10% offset because you will get booked.

[00:30:07]>> Yes. 10% is you don't need to pay any.

[00:30:11]>> Oh really?

[00:30:11]>> Yes.

[00:30:12]>> Up to 10% in in China.

[00:30:14]>> Okay. Okay.

[00:30:18]>> Almost road.

[00:30:20]>> I can pretty much guarantee if you were doing a well I mean not that we can do 130 but our our max is 110. So, if you were doing 121, >> oh, like 120. If you drive lower than 132, maybe you don't need to pay.

[00:30:37]>> Yeah. Okay.

[00:30:42]>> Back home, >> only warning.

[00:30:44]>> Yeah. Back home, if you're doing if you get booked by an average speed camera that's 110 and you were doing an average of 121, you most certainly will be booked.

[00:30:54]>> Do you agree?

[00:30:54]>> I think so. I mean, I don't know exactly what the tolerance is.

[00:30:57]>> I'm >> That's not public knowledge.

[00:30:58]>> I don't think it's 10%.

[00:31:00]>> Yeah, I doubt it.

[00:31:03]>> I believe someone posted on Reddit recently, actually, that that the tolerance is actually pretty low.

[00:31:10]You want to troll through Reddit, you might be able to find that post.

[00:31:14]>> Now, we're going pretty quick now.

[00:31:17]>> I feel like all airport entrances kind of start to look the same.

[00:31:21]>> Yeah, now we're crossing 100.

[00:31:23]>> Yep. I feel like we're going so fast.

[00:31:27]>> Is it going to slow down?

[00:31:29]>> Should I slow down?

[00:31:31]>> Okay, I'm just taking it off. I feel like I'm going too quick.

[00:31:33]>> Yeah, >> there you go. So, you've intervened there.

[00:31:39]>> Second intervention.

[00:31:40]>> I was going way too fast.

[00:31:53]trans.

[00:31:57]>> I think that was using a voice command to close the window, which is pretty cool.

[00:32:01]>> Okay, I'm going to activate VA again.

[00:32:04]>> Yes.

[00:32:05]Okay, off to the toe.

[00:32:23]One thing the um the Tesla had that this doesn't was that as a passenger here, I could look at this um screen here and it'll tell me how long I have to my destination, >> which I can no longer see. I wonder whether it is an option.

[00:32:39]>> It probably is.

[00:32:40]>> Maybe. Let's see if I can >> I don't know. I I don't want to dare go through all of this.

[00:32:50]>> This is good though. Planning well ahead. Saw a green line all the way from the toll booth that we had to take at the exit.

[00:33:00]And looks like we're approaching the airport very shortly.

[00:33:15]Another merge. This should be interesting.

[00:33:24]These merge points always fill me with thread in a foreign country.

[00:33:28]>> Yeah.

[00:33:30]Oh, there's a bus.

[00:33:32]>> Bus. Okay.

[00:33:35]Did this person get a summon?

[00:33:38]>> Maybe broke.

[00:33:40]>> Maybe.

[00:33:41]>> Crash.

[00:33:41]>> Oh, crash.

[00:33:45]Interesting.

[00:33:52]See that little um >> safety triangle?

[00:33:55]>> Yes. Yes.

[00:33:57]>> Right.

[00:33:58]>> Yeah.

[00:34:03]>> Cute.

[00:34:03]>> Yeah.

[00:34:13]Looks like we need to merge out eventually. That wasn't us. Someone else got >> peeped.

[00:34:32]One more cross. Yep. One more cross.

[00:34:40]I guess that's the other thing. As a human, I would only ever really dare like changing one lane at a time, but it can read all the lanes. So, we can actually know that not just this lane's clear, but the next lane over is also clear, isn't it?

[00:34:57]>> Yeah. With confidence.

[00:34:58]>> Yeah. Yeah.

[00:35:06]I so wish we had the traffic light countdowns.

[00:35:09]>> Yeah, I know.

[00:35:10]>> It's so good.

[00:35:14]>> Would it make people pay more attention >> or get more impatient?

[00:35:29]Interesting. It didn't indicate there.

[00:35:32]>> Uh, well, it doesn't really need to, is it?

[00:35:34]>> Must have heard me.

[00:35:36]>> There's a car behind us.

[00:35:37]>> Yeah, >> it actually didn't follow the lane it's supposed to be in, I think. Anyway, so we'll reach our destination in 5 minutes.

[00:36:08]>> So I reckon we should probably summarize at this point. M.

[00:36:12]>> So for me personally, things that does well, the planning, the route planning, >> all this green and yellow and red, knowing which lane it's supposed to be in, that's for me a big thumbs up.

[00:36:26]Yeah. I mean, it's honestly, it's hard for me to comment because these are conditions that we don't really have back home. Um like I feel like it's more willing to bend the rules because it's okay to bend the rules here if that makes sense.

[00:36:44]>> Um >> well the tolerance is greater. Yeah.

[00:36:46]>> Yeah. Yeah. Because it's it's also expecting other people to bend the rules as well. So, um, it kind of knows that other cars are also not going to be following the rules and therefore it also doesn't need to follow follow as as you were saying kind of like the lane keeping and all all that kind of stuff like being in this lane or that lane because sometimes there's like from one part of an intersection to another part of the intersection there's like two lanes that just becomes one lane after the intersection. So, has to be able to deal with something like that.

[00:37:17]Um, it may or may not have emergency intervened when you emergency interveneed. Hard to know.

[00:37:27]>> But on your point about um, yeah, it's hard to compare scientifically. I mean, it's not quite, you know, the same. So, >> we're comparing apples and oranges, apples and apples.

[00:37:35]>> It's got millions of data, millions of miles of data here in China, >> but not enough overseas. So to really be truly compar prep compared, we need the same amount of miles with Tesla FSD eventually. Yeah.

[00:37:46]>> In Australia, which >> yeah, until that point, we can't really say for sure.

[00:37:52]It's the same or better.

[00:37:55]>> But in terms of the way it accelerates and breaks, it's >> it's very good.

[00:38:01]>> Yep. Exactly. From the what we've seen so far here in China, it's very promising. You can and you can certainly tell when like when Tom took over, I could feel the difference in the driving. It was a lot jerkier and not as smooth.

[00:38:17]>> Yeah. But again, just to point out, this generationally better than um XGP. M >> but yeah, once again, looking forward to testing this in Australia one day.

[00:38:36]>> Yeah, that would be very cool.

[00:38:38]>> Especially if there's no subscription fee.

[00:38:41]>> Yes, >> that's what they're promising.

[00:38:56]There's a car stopped on the side of the road. I think it one more across.

[00:39:04]>> See what's it what's it doing now? It >> needs to get into that lane.

[00:39:10]>> So, we're coming to T3.

[00:39:12]Yep.

[00:39:22]It does helpfully tell you when it's going to or when you need to take over.

[00:39:25]It gives you a warning.

[00:39:29]>> Again, I think Tesla FC FSD doesn't do that.

[00:39:33]Doesn't give you warning.

[00:39:34]>> Now, interestingly, according to the navigation, we're in the wrong lane, aren't we?

[00:39:38]>> Eventually, cuz well, doesn't want to break the rule because it's a broken unbroken line. H wonder whether I should take over.

[00:39:48]>> You need to park near 62.

[00:39:51]>> Okay.

[00:39:56]>> I might just take over now. Getting pretty close.

[00:40:00]>> Okay.

[00:40:03]>> All right. We will stop the video then.

[00:40:04]>> Thanks for watching everyone.

[00:40:05]>> Yeah, we better pay full attention here.