Tesla Robotaxi - The Road to Autonomy

Transcript: We Rode in Tesla’s Robotaxi, Here’s Our Experience

Executive Summary

The Autonomy Markets team traveled to Austin, Texas, for an exclusive three-day experience with Tesla’s new Robotaxi service. During their extensive testing, they found the Robotaxi’s performance to be flawless and impressively smooth, even more so than competing Waymo rides. The vehicle demonstrated remarkable learning capabilities, adapting its pickup and drop-off points for better convenience over time, and confidently handled complex driving scenarios without a single issue, suggesting it’s ready to scale.

Key Topics & Timestamps

[01:21] First Robotaxi Impressions

Grayson Brulte and Walter Piecyk share their initial reactions just minutes into their first ride, noting the novelty of being in a car with no one in the driver’s seat while moving through Austin traffic.

[02:27] Smoother Than Waymo? 

A direct comparison of the ride quality reveals the Tesla Robotaxi is smoother, less aggressive on turns, and offers a more comfortable ride than Waymo’s Jaguar vehicles. Grayson and Walter note its braking is more predictive and less “herky jerky”.

[05:01] Performance & Scalability

After numerous trips with zero issues or even minor glitches, the hosts declare the Robotaxi’s performance a “10 out of 10” and agree that the service seems ready to scale to more users and a larger area.

[09:59] Real-World AI Learning

Grayson and Walter describe how the Robotaxi adapted over two days; initially, it made them walk a block from their hotel, but it later learned to navigate the hotel’s difficult circular driveway for direct pickups—a task Waymo never accomplished.

[12:21] ODD Expansion

Grayson and Walter spot Tesla vehicles at a supercharger that appear to be calibrating equipment for service area expansion. Their speculation is confirmed when Elon Musk tweets that the Austin operational design domain (ODD) is officially expanding.

[15:40] Austin’s AV Ecosystem

The hosts discuss the overwhelming presence of Waymo vehicles in Austin, estimating their fleet has grown to around 200 cars. They note the beginning of a direct competition between Tesla and Waymo in the city.

[19:05] The Robotaxi App

The hosts explain that the service uses a separate app called Robotaxi which is seamlessly linked to the user’s main Tesla account, pre-loading their profile and credit card information for a frictionless experience.

[21:40] The Grocery Store Test

An impromptu test reveals a key performance difference: the Tesla Robotaxi executes a safe, curbside pickup at a supermarket, while the Waymo stops inconveniently in the middle of the busy parking lot, blocking other cars.

[26:47] Why Austin?

Grayson and Walter debate why Austin has become the central hub for AV development, concluding it’s due to a combination of a favorable regulatory environment and, most importantly, the city’s ability to attract the engineering talent that drives the industry.

[30:25] Blending In vs. Standing Out

A key observation is that the Tesla Robotaxi, looking like a normal Model Y, blends into traffic unnoticed by other drivers. In contrast, the sensor-heavy Waymo vehicles attract significant attention and are treated like a “science project” or a “Disney ride”.

[33:18] The Safety Attendant

Grayson and Walter observe that the safety attendant’s screen has no fingerprints after dozens of rides, suggesting they aren’t intervening via the screen. They also confirm the service operates effectively in the rain, debunking speculation that it would shut down.

Subscribe to This Week in The Autonomy Economy™

Join institutional investors and industry leaders who read This Week in The Autonomy Economy every Sunday. Each edition delivers exclusive insight and commentary on the autonomy economy, helping you stay ahead of what’s next. 

Watch the Full Episode of Autonomy Markets

Full Episode Transcript

First Robotaxi Impressions: A “Flawless” and Smooth Ride

Grayson Brulte: Walter, it’s reveal day on autonomy markets. You know where we went? I know where we went. To our listeners and viewers, Walter and I headed to Austin, Texas. Unfortunately no cowboy boost for Walter and no cowboy hat for me. But boy, we had a great time. We were invited to go in Robotaxi. Wow, wow, wow. It really impressed Tesla. Thank you kindly for the invitation. You’ve exceeded our expectations. This episode of Autonomy Markets is dedicated to Tesla, Robotaxi. And in the coming weeks we will share more in depth stuff of the other company’s vision, the other technology we saw. But for this week, it’s Tesla Robotaxi, special on autonomy markets. 

Walter Piecyk: Yeah, that’s exactly right, Grayson. It was, it was quite a trip. Um, it’s gonna give us content, I think, for, for another couple episodes for sure. And we’ll share more videos over time as, as well as, um, some more information on, on kind of what transpired. The way we’re gonna do this is, is we recorded a lot of this stuff in the back of the Robotaxi itself, which gave us, you know, a flawless trips in order to do that. Uh, and in this first segment, you know, we we’re, we always have to show the first ride. Don’t worry, there’s not gonna be lots of videos of just us, you know, looking out the, the, the window. I’m sure you’ve seen plenty of those from two weeks ago. We’ll give you our first impressions, oppressions after a day. And then impressions after the three days of, of Robotaxi, uh, trips, uh, that we took. So without further ado, let’s go to this, this first reel of, of Clip First Ride. 

Grayson Brulte: Let’s go see First ride. Yeah. All right. And we’re good. Alright, thanks. Okay, Grayson. Then we’re a couple minutes in to our first ride. Yep. Without trying to sound like too much of a. Tesla fanboy. Uhoh. I can call Tesla Stan. Now Tesla. Stan, what is your initial your initial reaction? They did it. They did it. They did it. I mean, the bottom line is we’re in a car. There’s not anyone in the driver’s seat. No. It’s, you know, progressing through the streets of Austin with a modern amount of traffic. 

Walter Piecyk: There’s no, yeah, there is no driver in the driver’s seat. What’s your overall impression in terms of the car’s feel smooth? 

Grayson Brulte: I find it smoother than the version of FSD that I’m running at home. I find it when it comes to stoplights, when it comes to stop signs it, I don’t wanna say perception ’cause I don’t have data to back this up, but I feel like it’s looking further ahead and it’s adjusting to coming to stop signs and stop lights further out than my version of FSD, which I’m currently running. 

Walter Piecyk: And I gotta say for me, as our, um, listeners may recall, I kind of salted some of the early riders from two weeks ago who, you know, vary with a lot of hyperbole, saying this was much better than Waymo. I still don’t think this is much better than Waymo, but since we have been kind of testing back and forth between Waymo’s. And the Tesla robot taxii. It is, I think, an overall smoother ride, as you said, as you’re coming up to stop stoplights, which we’re doing now, it’s just a much smoother stop. It’s a little bit less aggressive around the turns. It’s just overall a more comfortable ride. Part of that, by the way, is the Y itself is just more comfortable than getting crammed into a Jaguar, which again, when compared to a Y, it’s just a much better experience. 

Grayson Brulte: I like the seats in this vehicle, find it very comfortable Passeng perspective, but we just came to a stop. The vehicle about six to 800 feet before started to to slow down, came to a complete stop and then slowly moved up. But we didn’t feel there was none. None of that herky jerky was very, was very smooth, as I would say, a highly trained professional driver and delivery service would drive very similar to that, not a robax necessarily. So well done Tesla. 

Walter Piecyk: And at the end of the day, I’m not sure it totally matters, like if both of them. Feel generally pretty good. I feel like this is almost like when people were saying, what’s the difference between 4K and hd, unless you’re in the Best Buy. Seeing them right next to each other, like we’re effectively doing today, going back and forth between Waymo’s and robotaxis, so you can really tell the difference, but if you’re just taking a Waymo or you’re just taking Robo Taxi, they just both feel obviously very human and normal. 

Grayson Brulte: In terms of the comfort of the ride, do you know what the comfort of that ride does? It builds trust. When you put an individual in here and they experience how comfortable it is and how smooth it is, it’s gonna help build public trust. 

Walter Piecyk: Yeah. It just, you again, like we’ve all experienced, I think with early Waymo rides, you just get comfortable. It’s, it’s novel at first, not seeing anyone behind the driver’s seat, and then you go by your business looking back at your phone, um, you know, because you’re just trusting what’s going on in the car. And I think that’s ultimately what’s gonna happen with the Robotaxis. 

Grayson Brulte: I want to point out, because cer very few people knew that we were coming and when I got in the vehicle. I never felt nervous once I felt comfortable from the moment the vehicle took off. And that says a lot. 

Walter Piecyk: I think there that is part of the comfort of the ride, meaning like the herky jerkiness, you’re like, oh, if it’s like making a sudden turn, or like it breaks a little bit like some of the earlier versions of FSD did, it does make you nervous. Coming up to situations, but when you have a more comfortable ride, it reduces. I think that that nervousness about allowing the car to do what it’s gotta do. 

Grayson Brulte: I came here uncertain what to expect, and I’m leaving here completely impressed with the performance of the vehicle. 

Walter Piecyk: I mean, the bottom line is we’ve had zero issues, like not even a short break, you know, like sometimes you have an FSD where whether it’s a shadow or whatever it is, like zero. Issues with this software, um, load. It just seems, you know, it seems ready, it’s ready to scale. 

Grayson Brulte: That’s what I would say. I would say it’s truly ready to scale. And we got the tweet from Mr. Musk last night that they’re gonna expand the ODD, so there’s extreme confidence in the way it’s performing. 

Walter Piecyk: Right? So I guess, you know, now we’ve already, we talked about this maybe in other segments, in that they’ve hit the next milestone, which is increasing the ODD. In a matter of a couple of weeks, assuming that it does happen this weekend, as, as Elon effectively promised, next thing to do is add cars. Like unlike other autonomy car companies, you don’t, you’re not gonna have people sitting there wiring up additional sensors. They should just, you know, provision a new Y. The only thing they have to do is put the Roboto Taxii label on the side of the car, right, to add some volume here and get some additional safety attendants. 

Grayson Brulte: Exactly. If you look at it from a maintenance perspective or a charging perspective, there’s plenty of superchargers all over Austin, but there’s nothing from a maintenance technical supercharging perspective that would slow down the expansion of Robax. 

Walter Piecyk: I mean, clearly the number of rides we did, which were a lot over the last three days, are not as they would say statistically significant. But I think, you know, they’re certainly at a point where I’m guessing they’re gonna be adding more and more people, um, into the app. And, you know, given the focus on this company, if there’s even the minor glitch, I’m sure we’ll see it on YouTube or Reddit or wherever. So I’m not as concerned about that. I think we were more here to get a feel for the ride, which we’ve discussed in the past segments. And, but I agree with you. Like I, I am, you know, very surprised that we didn’t even have the slightest glitch in any of the rides that we did. 

Grayson Brulte: Yeah, it’s been a 10 out of 10 or a hundred out of a hundred. It, it’s, it’s truly, truly performed and it’s, it, as I said, it has, it completely exceeded my expectations. And we said this on a past podcast, Waymo, you have a real competitor now. Yeah, it’s Tesla. 

Walter Piecyk: And it was a more comfortable ride than Waymo, which I acknowledge and you know. For our listeners, I will be now picking up my new why this Sunday, which I saw I now will have hardware for and getting the latest FSD software. So it’ll be good to be able to compare this in a couple of days to what I’ve experienced here in this market. 

Grayson Brulte: Walter, those were great clips. Fully autonomous, no interventions. Tesla performed. They beat my expectations. Wow, wow, wow. What’s next? 

Head-to-Head: Robotaxi Learning vs. Waymo’s Performance

Walter Piecyk: In this next set of clips that we have? We thought it was important to kind of show. How the Tesla, you know, pulled into some areas that it then had a back out of and doing three point turns. And then more importantly, how the Tesla, we, even in the three days that we were there, seemed to learn on day one. In the hotel that you and I were staying at, our pickup point was like a block away. You’ll see that in the video. But by day two, by the end of day two, the Tesla was actually coming into the hotel in the little mini circler, which you’ll see on the video. Um, in order to pick us up and, you know, just to be clear, not that, you know, the Waymo was never able to accomplish that. ’cause we did take a lot of Waymo’s and Teslas while we were there, and the Waymo never did accomplish that task of coming into the hotel. So, without further ado, let’s hit the second group of clips. That’s really from the app. 

Grayson Brulte: So we’ve discovered that our Tesla is learning at our hotel is a tricky entrance. Mm-hmm. For both Teslas and Waymo’s. They, you know, all the last two days we’ve been gotten getting picked up on the streets and dropped off on the streets. And then what happened tonight? It went into the valet zone first time and the Waymo can’t do that yet. 

Walter Piecyk: Yeah, I mean, the Waymo hasn’t done it yet. Whether it can or it can’t, who knows? But, um, I don’t know. Definitely. It’s just so odd because all day it was dropping us off, off in the street and we had to walk a little bit, uh, for the pickup. 

Grayson Brulte: It was amazing to watch the vehicle learn. It’s like watching a child go from walking to running. And it wasn’t just the, the situation at the hotel. We also had another situation, Walter, as we were. Doing our grocery store experiment to see what would happen. The vehicle went on a turn to the left as it was supposed to, but the vehicle determined it couldn’t legally cross the road, so it goes back around. And then what does it do? It goes to the light the next time. And when in that entrance, it was truly amazing to watch the vehicle learn over the time that we spent the vehicle. 

Walter Piecyk: What’s next? I mean, first of all, on that point, you, you’re a hundred percent right. I, I thought we were gonna get stuck in endless loop of it trying and trying and trying, but it just, rather than making the first left, it basically made the second left, uh, to complete the, the task that was impressive. 

Scaling Up: Elon Musk Expands Austin ODD Amid Market Competition

Walter Piecyk: So in our next group of videos. Um, you know, we, we were, came upon at one point at a supercharger, a couple of Teslas that looked like they were preparing for an increased ODD, and that was clearly one of our takeaways that things were going so well. It was just a matter of time. I. Um, that the ODD expanded. So we commented on that a little bit and thankfully while we were there doing all these test rods, obviously Elon tweeted out that in fact Austin will be increasing. So let’s go to those, those clips now. 

Grayson Brulte: Walter, we’re having dinner all suddenly phone start buzzing. Elon’s tweeting. No, he is not beefing with the president. No, it’s not. The American party is hosting their. Independence Day. It’s Elon’s expanding the Austin ODD for Robo Taxii. What do you think? 

Walter Piecyk: I mean, it made sense. We, you know, we’re almost on day three. We’ve had a lot of rides, we’ve had no issues and I. Uh, we’ve both been talking about, you know, we just need more cars and, you know, a larger ODDI feel like it’s, it’s ready. I mean, I think you keep the safety attendant here, they can scale up and just continue to, to show out the service. Um, and it also makes sense because we saw earlier today a bunch of Teslas with them with. You know, in fact, LIDAR sensors on them that I think they were using to try and expand and look for new areas to expand. Um, you know, a couple of them at a, at a charging, which I think we’ll post some videos and pictures of later. 

Grayson Brulte: Yeah, it’s a, it’s a positive step. It’s a reinforcement that Tesla. Believes that the technology is running to truly scale Robotaxis. 

Walter Piecyk: Yeah. This is just one of the, the common themes that you and I have discussed over the last two days. It’s like, this is great. It’s working great. Now it’s time to expand. It’s only been two weeks, so don’t wanna put the cart before the horse, but things seem to be going well enough that, you know, maybe it’s time to, to move forward no matter what it is. 

Grayson Brulte: Walter, an expansion is an expansion. The expanding ODD just validates robax unit. Its technical capabilities. We’re looking forward to seeing where they expand. Next. A little pitch and a plug for me. Please come to Florida. We have a great regulatory environment and would love you here. 

Walter Piecyk: So for our next group, um, it was hard not to notice just all the, the tons of Waymo’s there. So we, we chatted a little bit about that from the back of a Waymo, um, and also related something we found new in the Uber app, um, that you were required to use, um, in order to get a Waymo. And while we’re talking about apps, you know, I think race and you have a video here that broke down a little bit about how. Um, the Tesla app itself, uh, was performed what a market for autonomy, seeing Zoox and AV ride and, and Volkswagen and, you know, Waymo’s and Tesla. But the one that’s kind of dominant, even though we’re seeing autonomous cars everywhere, is Waymo. When this thing, when Waymo first launched with Uber in this market, you, it was limited based on car volumes to whatever I, I think we said 35 or 40. What do you think that number’s up to now? 

Grayson Brulte: I would say it has to be around 200 or so. And we were having breakfast yesterday. I looked out the window and I counted 13 Waymo’s and it seemed You turn left, you turn right. You see a Waymo? They’re everywhere. 

Walter Piecyk: Yeah, it is. It is crazy how many they’re seeing. I think two hundred’s, probably around the right number. Probably room for expansion even more. Um, gotta be hurting the UberX driver out there. Or in my case, Uber comfort. ’cause we were getting Waymo, um, you know, when I was selecting Uber Comfort. So it’s gotta be. It’s starting to impact, at least in this market, the human driver, given how, how, how many times you can easily get them. The only time that we’ve had a human driver is when we’re outside of the ODD. Every other time we’ve been paired with a Waymo consistently. 

Grayson Brulte: Yeah. So I would guess from, you know, we know that that Elon’s expanding the ODD here for. Tesla, I would guess that within a couple of weeks you’ll see also Waymo, given the number of cars that they have also expand, you know, their ODD in this market. And then it becomes interesting. It becomes a competition per se, of who’s gonna expand faster and who’s gonna offer rides. And then when do we get pricing involved. When do we start have a pricing war? 

Walter Piecyk: Once the ods are very similar in size. I think we’re so far off from that, given the limited number of overall cars, and given the fact that any of these autonomous cars are a higher quality than a human driven car, more consistent, cleaner. More consistent ride. So I don’t think you’re, you’re gonna see a price for anytime soon in, in, uh, in the world of autonomy. If anything, it’s the opposite, right? Uber comfort is they’re gonna drive that. When you have a separate Waymo, you know, uh, category that a hundred percent should and will be, I think, priced above where UberX is. 

Grayson Brulte: It’s gonna be very interesting to watch, see how the, the market dynamics work out. I tell you, we’re gonna keep our eye on the Austin market in my rush to book a Waymo. I actually hit business comfort and in my settings I had changed it to we to Waymo default, and they’re now allocating me Waymo’s in Business Comfort, which obviously probably has a higher price than, than uh, UberX. I think you might’ve been overcharged maybe. 

Walter Piecyk: But I’m happy to have gotten a Waymo, frankly, rather than a human driven business comfort because of, again, the consistency of this ride, how it feels to cleanliness and even the temperature versus some of the human driven, uh, Uber rides that we’ve all had to experience. 

Grayson Brulte: The funny thing though, every time we get into Waymo 69 degrees, every time we get into a Tesla 66, is that Waymo having a little fun? I don’t know about that. 

Walter Piecyk: I, I think it is the thing that, another thing about the Uber app, from my perspective, the anxiety and the stress. Are we gonna get to Waymo or are we not gonna get to Waymo? Are we gonna get to Waymo? We’re not gonna get Waymo. Little bit of stress there. Well, it’s only stress ’cause we’re very focused on autonomy and want to get those cars. But yeah, it does go back to, I’d love to be able to just not hope for a Waymo, but specifically select a Waymo. And to be fair, the reality is it’s we’ve, other than one location we were in, the allocation of a Waymo to us has been pretty good. And we’ve seen a ton of Waymo’s in the market. And you’ll see in one of our other segments, the massive depot that they now have here in Austin for Waymo’s. 

Grayson Brulte: Andrew McDonald. Sir, you tweeted me on X That a robot taxi tiers coming after this time in Austin. Sir, I think it’s time to accelerate it ’cause you got something special here. And don’t forget to charge more because this is a much better ride than any Uber X. Even though wallet has to pay more. 

Walter Piecyk: I agree. This is a luxury product. We’re in Austin, Grayson, you know. How are you? Is this app part of the Tesla, your Tesla app or is it a separate app? 

Grayson Brulte: It’s a separate app called Robax, but it’s linked to my Tesla account. My settings, my profiles for my Tesla Model three at home. Go in here. My account with my credit card on file, Tesla is in Robax. Two separate apps, but one master account, which raises the question Walter, does that accelerate the adoption of Tesla owners to using Robax since it’s all fully integrated into your account? 

Walter Piecyk: I mean, it definitely reduces friction. Anything that reduces friction in the world of payment, you know, should enable that. It’s not like you have, I mean, you still have to download another app, so I think at some point they’d have to integrate it into the app, which I keep open. As a Tesla owner, so if they put it in the Tesla app, but even if it doesn’t, to your point, if billing is connected, you know, reducing friction is always positive for adoption. 

Grayson Brulte: because I didn’t have to go through the hassle of getting out my credit card, putting in, I put my email address in and logged in my tests, the credentials, and then boom, away it went. 

Walter Piecyk: I mean that’s true, but at the end of the day, this is gonna be a mass market product. Uh, product. So if you just, if you look at the sub subsegment of the market, which is Tesla owners, I’m not sure that that’s gonna be the swing factor one way or another of, you know, this thing getting to scale. 

Grayson Brulte: but I’ll give you a little pushback. Bestselling car in the world. Test the model wide. That’s the car we’re in now. 

Walter Piecyk: Sure. Well, we’ll see. We’ll see. I mean, anything is, it’s incrementally positive. I just don’t agree that it’s a dramatic. Difference in how they can get to scale. 

Grayson Brulte: I disagree. And that’s what makes this podcast fun. Exactly. Robax app is great. It’s fully integrated in my Tesla account to Seamless experience. What I’m watching forward is Robax app eventually emerge into the Tesla app. Walter, what are the next clips? 

Walter Piecyk: I mean, we already talked in the first segment in the clips at least of, of how we felt the, the Tesla was a smoother ride than, than the Waymo. In this next group, we, we saw some difference in the performance. Of how these Waymo cars pulled up to grocery store, um, you know, parking lots versus the Tesla, and also what type of roads they took. So a little bit comparison discussion between. Waymo’s and Teslas in Austin. One of the things that I’ve already noticed, ’cause we have been kind of swapping back and forth, uh, with Waymo rides, is it seems like, um, so far, at least Teslas like to take main roads and Waymo is taking a lot of back roads. What do you think that means, if anything? 

Grayson Brulte: Could it be a speed issue? The Teslas are, they’re more confident, driving faster than the Waymo’s. To me, there’s, it’s really a routing issue, which goes to who’s routing the algo for the Waymo, since it operates Uber, is it an Uber routing algorithm or is it a waying Waymo routing algorithm, or is it a speed issue? The Tesla clearly has no problem driving at higher speeds here in Austin. That it’s something I’m very curious about. 

Walter Piecyk: I mean, we were in a Waymo that I think was traveling 50 miles an hour. It was adjacent to the highway, so it was certainly capable of doing higher speeds on that road. But I think you’re right, their algo seems to be sending us more in the back roads, and it reminds me of when I was in la and in that case, it was clearly staying away from the highways, which made what would’ve been a 20 minute trip. A 40 minute trip staying off the highway. Um, but it seems to apply that here. Not staying off highways, but again, staying off main strips. We, this might just be our limited sample size, so it’s hard to say we’ll be taking more Waymo’s over the course of the next two days. But, um, it was an interesting observation. 

Grayson Brulte: but let’s get another data point. We have taken human driven Uber since we’ve been here, and the routing has been odd, to say the least, with a lot of back rows versus main roads. So. It’s another data point that perhaps it’s the Uber routing algorithm. 

Walter Piecyk: Yeah, it’s, it’s definitely impossible there. I know I’ve taken Ubers, you know, back in New York where sometimes they set it to avoid the tolls and they take, you know, a circuitous route as opposed to a faster route that costs them money. But, um, I don’t think that they would set the wing up to do that, but yes, you’re right. It, we could, it could just be a Uber issue, um, as opposed to a Waymo issue ’cause it. Sending it the location, although I think we just met with a someone that in autonomy that was saying, typically what happens if Uber has a partner is they send the location and the algo of the autonomous company will determine what the route is. So. While you’re right g Grayson, it may be Uber. I’m, I still think it’s, it’s a Waymo decision to do that to the Waymo viewers. 

Grayson Brulte: Right in Let Wall, and I know. Does Uber set the algorithm in Austin or does you Waymo set the algorithm? We’d love to know. So we’re on our way back to a supermarket and we’ve kind of discovered something interesting by accident. First we took our, um, Tesla to a supermarket. ’cause we were asked like, that’s, that’s been challenged. That’s been a challenge for other autonomy companies in terms of how they drop people off at supermarkets. Um, and we went right to the curb and then when we got there, um, I was in, in a bit of a rush. Walter got a phone call ’cause he wrote the most red. Paper on the Bloomberg terminal. Should Tim Cook resign as CEO of Apple? My opinion, time for a product guy, Walter put it out there and guess what? Ring, ring, ring. Hello? CNBC on went Walter, right? So I was a little worried about getting in time for the hit. Um, so I hit up a Waymo because when we got out of the car, the Tesla, for whatever reason, it doesn’t allocate you to the Tesla. You just get out of. We were gonna wait for 25 minutes. So my anxiety started kicking in about getting back in time for the, for the CNBC hit, and, you know, so I hit up a Waymo, which, which ultimately came basically at the same time as, as, as your Tesla. 

Walter Piecyk: It did, but as a viewer, Chuck Price pointed out there is a big difference. And guess what we found out? The Waymo picked Walter up in the middle of the parking lot blocking the cart lady when Mrs. Cart can’t go in there and get it, you blocked it. Yeah, I was blocking a couple of different cars, so it’s kinda a weird thing. I’m not sure why Waymo’s doing that. It’s definitely safer to go all the way to the curb than in the middle of the parking lot, but we’re gonna test it again. We’re headed to another supermarket and we’re gonna call up a Waymo and see where it, where it tries to pick us up. 

Grayson Brulte: And do you know where the Tesla picked us up? Curbside. The grocery store test. That’s an interesting test because it’s something that we do every day. Chuck Price, thank you so much for pointing it out to us and for giving us the inspiration idea to do it. 

Walter Piecyk: Walter, what’s next? So in this next group, we kind of put a couple together. First we talked about like, why is this even occurring in Austin, uh, as opposed to other markets. Um, we talked about, you know, we thought long term with private ownership because this could be a huge opportunity for private ownership. So when we saw these sensors and we actually saw a depot where sensors were maintained, what that might mean in the long term. Uh, and then finally it’s always a popular topic. The attendant, the safety attendant, that was, that was with us on all these trips for. We have some, we have some EC clip, uh, commenting about our attendant, the to Tesla. 

Grayson Brulte: The attendants are fantastic. To the listeners of viewers who have not had the chance to go into a robotaxis say, how are you today? What can you tell me? Visit  tesla.com/robotaxi. Taxi. Let’s roll the clips. A lot of people believe that we’re here in Austin, ’cause Texas is open for business regulatory wise, and then Austin’s a city where you get a good, diverse group of people willing to live. But you know, Florida’s got those areas as well, Southern Florida. Why, why, why is everyone congregated here in Austin? There’s so many autonomy companies here and not in Florida. 

Walter Piecyk: The engineers want to be here is the bottom line is closer to California, but the engineers want to be here. The engineers are driving this factor of all the work that I’ve done over the last decade in Florida. It’s been very hard to get engineers to relocate there, but we do have a breakthrough. Waymo will be coming live online next year in Miami zoo’s at who knows when they’ll be coming online. But it really comes down to engineers have been driving it, not necessarily the policy individuals. 

Grayson Brulte: But why would an engineer not wanna live near Miami? Which seems like a pretty fun city to live near. Not that Austin isn’t, but Miami also is as a place to attract talent, perhaps better job hunting. 

Walter Piecyk: I think Miami’s a lot of fun. I know Miami’s a lot of fun, frankly. I. It, maybe it was just a preference. I have no idea what attracts engineers here. Maybe it’s the cluster of engineers that are already here. Were in Miami. It’s not necessarily that cluster. I mean, there has been a lot of venture guys over COVID that were trying to attract people to, um, Miami. But the reality is Austin has had a, I think a long history of technology companies that are here. So, you know, I think that’s probably contributes to it as well. 

Grayson Brulte: It contributes to it. And then the other thing here in Austin, there’s an ecosystem as you know. VW is testing, getting area to deploy here. Zoox is testing, getting ready to deploy here. Tesla has deployed here. Waymo has deployed here. So there’s a, there’s a long history and there’s a really great regulatory environment to deploy it as well. 

Walter Piecyk: And it from an infras, from a manufacturing standpoint too, there’s, there’s definitely a history of manufacturing that exists, exists more here than perhaps in Florida, anywhere in Florida. You know, another thing that occurred to me in taking these rides. And just again, noticing how people don’t really notice us in these robotaxis. ’cause there’s not a ton of sensors on. There is, you know, as we get five, 10 years down the road into private ownership, the maintenance of a lot of these sensitive, um, lidars and things that are on competing vehicles. And how that might restrict or maybe create a whole new cottage industry but restrict, you know, willingness to purchase, um, cars that have to have sensors that might need to get recalibrated in services. I mean, we saw a massive service demo depot that Waymo has. Um, with a VO, um, I don’t know. What are your thoughts? 

Grayson Brulte: I don’t know how you’re gonna be able to scale a personal autonomous vehicle with lidar and radar because the way the Tesla is built is a camera vision-based system. It’s a simple software update to fix, so the maintenance costs. Go down tremendously. 

Walter Piecyk: I mean, the answer is you make more durable, um, sensors. You don’t have spinning lidar, you have solid state lidar. I think what you’ll probably see with Waymo’s going forward is fewer and fewer sensors. I’m not gonna say that they’ll get rid of lidar, but maybe fewer lidars and less. Sensors to have to maintain. But um, it’s something to consider when you think about longer term adoption of autonomous privately owned vehicles. 

Grayson Brulte: And then you also have to think about the service aspect. How are they gonna be serviced when they have to go in for service? 

Walter Piecyk: Yeah, for sure. So anyway, just another, another tangent on. Thinking, spending multiple days in the hotbed of autonomy right now. Austin, Texas. Yeehaw. Well, that’s it. We’re not gonna inundate you with a ton of clips, at least in, in week one. That’s all we’re gonna share with you this week. But, um, before we we’re, before we go out with our final recorded clip, Grayson, is there anything else you wanted to add that, that, uh, that we didn’t hit today? 

Grayson Brulte: The one thing I’d like to add is that the Robotaxis blended into the environment. Nobody notices us except one gentleman. And you know what, that gentleman was driving a cyber truck. We got a big smile. A, a thumbs up. Keep going. Rock and rolling. One cyber truck driver. Everybody else around Austin didn’t even notice that we were there. We just simply blended in and we pulled up the stop lights. People would not look over when we got out of the cars with all of our gear, the GoPros, the, the mics and everything, no one really asked us. Meanwhile, still these Waymo’s, even though there’s tons of these Waymo’s in the market. They’re still being treated like a science project, a, a Disney ride, and they, they attract a lot of attention because of all of the sensors. 

Walter Piecyk: And, you know, it kind of reminds me of, of, of stuff that we heard from Uber in terms of Austin, in terms of expanding the market, the market’s expanded because there are tourist type, um, interest in driving these cars because of all these, these sensors. But I guess the question then is, what does that really mean going forward? If it has the car just kind of blends in with the rest of the car environment, it could be viewed as a strategic advantage. As it scales. 

Grayson Brulte: I mean, it’s possible. 

Walter Piecyk: It, it’s certainly, at the end of the day, what’s gonna grow this market is not necessarily, you know, being a tourist ride, being a Disney attraction ’cause, because as we’ve already seen in Austin, that only gets you to a certain amount. Um, what this industry is about is, is taking percentage of miles driven on the road from one or 2%, which is where it’s at today for ride share. To 10, 15, 20%, or just even for autonomy, whether privately owned or ride share. Um, so maybe utility is more important. And frankly, you know, at CES couple years ago, I sat on panels or watch panels where they were meeting with community leaders. And the community leaders were talking about, Hey, they’re gonna have these cars with these major, these sensors coming into our neighborhoods. We need to work together as a community. Um, so that everyone’s pleased with what’s going on. So maybe it’s something that. Facilitates easier adoption on a market by market basis. 

Grayson Brulte: That’s a very valid point. And I’ll say it could because model wise are part of a majority of the communities around the United States. It’s the bestselling vehicle in the world out outside of China minus BYD vehicle. So it’s just, it’s just another member of the community that potentially could accelerate that adoption and it could also help to build that trust. And when you get to personally own autonomous vehicles. I think that’s when you’re really gonna see the acceleration there. ’cause it’s not like, oh, the Jetson mobile showed up at Mr. Smith’s house. No. Oh, it’s the model Y. Oh, that’s cool. My wife has one. And so it starts to build that trust and build that adoption. 

Walter Piecyk: I mean, the counter argument to this is you want people to see these autonomous cars in the community so they get comfortable with them and they try and get in them and you know, you, you build adoption that way. So I don’t know. This is one of those things we’ve had great. Listener feedback in prior weeks. What are you the listeners think? Is it better for Tesla to blend in or is it better to have you know, big logos on them? You know, identify them as the robotaxis in the community. Let us know, you know, in the comments. 

Grayson Brulte: Let us know in the comments. And while you’re at it, let us know what you think of putting signs in the vehicles to identify ’em. What do you think there? Should there be emojis to communicate? People let us know that as well. 

Walter Piecyk: And as always, thanks for tuning in. 

Grayson Brulte: Thanks for tuning in, and we’ll just sign out with this last clip. See you next week. 

Walter Piecyk: See you next. 

Grayson Brulte: Here we are again. We’re back at it, Walter. We’re not dancing in the rain. We’re driving in the rain. 

Walter Piecyk: That is true. We’re now in reigning conditions and know the service hasn’t shut down here in Austin for Tesla’s robotaxis service. The other thing that we, me that, you know, I think we noticed, I, I think everyone saw in a lot of these initial videos, um, the attendant in the car, they have their hand on the door. Who knows whether it’s a kill switch or not. The other way for them to potentially stop the car is on that screen. You have pullover stop and lane and support. Um, and I think one of the things that we notice, Grayson, is I don’t see any fingerprints on the screen. And we know, we both know having Teslas, how easily it is to put fingerprints on there. 

Grayson Brulte: Dozens of rides. No smudges. It’s a clear sign. 

Walter Piecyk: I mean, it’s at least one sign. Again, you know, there is the kind of the hand on the door, which, which could be a trigger. I don’t know. We were talking earlier about how whether, um, maybe I. When the attendant finally leaves the car, they don’t want bad actors hopping outta the car mid ride and maybe the entire car is wired up, um, to basically if someone tries to open the door to stop the car in the lane to prevent any issues. But could be we, we will find out once. The safety intendant is removed at some point when that is, we don’t know, but we believe it will be soon. I also think it’s not that Tesla’s the first company to have the attendant in the passenger seat, but I think that you’ll see that as a rising trend among autonomy companies because. You know, not having someone behind the wheel is obviously very impressive for whoever happens to hop in the car. You still have someone there to, to effectively stop the car when you need to, but I think you’ll see more and more as, you know, companies go through their different models that, that someone sits in the at in the passenger seat, I’ll call it the tendency as opposed to behind the wheel. 

Grayson Brulte: I think it’s a, it’s a positive step and it also emphasizes safety that they’re doing this in a very diligent manner. 

Walter Piecyk: Yeah, I mean, that gets back to the whole safety model or the safety culture that we talked about before, but we don’t need to, to go back into that anyway. Uh, so we, you know, here we are in the rain. 

Grayson Brulte: No smudges on the screen. 

Walter Piecyk: No smudges on the screen. No smudges on the sensors. No lidars cars driving in the rain. Let’s just show the windshield wiper going. There we go. Just to prove that, I mean, it’s obviously a light rain, not exactly a super test, but I think there was some speculation that these things would stop in any type of rain. So we’re still going. 

Grayson Brulte: Still going, Walter, we took Autonomy Markets on the road. I just wanna say thank you. I had a ton of fun with you. It was great. 

Walter Piecyk: And this, and we have some additional road trip plans throughout the year. I mean, this is a new feature of autonomy markets, but yeah, that’s a wrap on Austin. What a what a what a great three days. 

Grayson Brulte: Look for more, look for more road trips from us. Look for weeks of content coming from markets, and be sure to tune in every Saturday at 10:00 AM on your favorite podcast platform. The future is bright. The future autonomous. The future is keep Austin autonomous. 

Key Autonomy Markets Episode Questions Answered

How does the Tesla Robotaxi ride comfort compare to Waymo? 

Grayson Brulte and Walter Piecyk found the Tesla Robotaxi to be an “overall smoother ride” than the Waymo. They described its braking for stoplights as less aggressive and its stopping as “much smoother,” leading to a more comfortable and less “herky jerky” experience overall. The Model Y vehicle itself was also considered more comfortable than the Jaguar used by Waymo.

Did the Tesla Robotaxi show any signs of learning or improving over time?

Yes, the vehicle demonstrated clear learning capabilities over the three days of testing. Initially, it would pick the hosts up about a block away from their hotel. However, by the end of the second day, it had learned to navigate into the hotel’s small circular valet area for pickup, a task the Waymo vehicle never accomplished during their tests.

What is the user experience like for the Robotaxi app? 

Robotaxi uses a separate app from the Tesla app, but it is linked to the user’s master Tesla account. This integration makes the process seamless by using the credit card already on file, which “reduces friction” and eliminates the need to manually enter payment details.

Now that you have read an Autonomy Markets transcript, discover how our market intelligence and strategic advisory services can empower your next move.