962 - Building An Autonomous Golf Cart Network W/ Danny Minutillo (Nav Robotics)

Mat Sherman (00:01.4)
Alright, we are live. Alive and breathing. Here we are with Danny from... It's just Nav or Ride Nav? Nav Robotics? Yeah. You can put anything, like blank robotics, sounds legit. Robotics is just like it makes your company name sound so freaking badass.

Danny (00:12.431)
nav robotics. Yeah. Right now just the domain. Yeah. Nav robotics is really expensive. So.

Danny (00:26.869)
Yeah, very high tech, right?

Mat Sherman (00:29.57)
But the thing is you're actually building robotics. You're doing the thing, so it makes sense. Cool. Well, excited to have you on the pod. So for people that are listening, Danny and I know each other fairly well in real life. So it deviate slightly from just talking about his company, Nav Robotics. But y'all don't know what he's working on yet, so let's just start there. Danny, what is Nav? What are you cranking on?

Danny (00:35.229)
Yeah, yeah, for sure.

Danny (00:57.353)
Yeah, yeah, for sure. Yeah, so excited to be here as well. Yeah, so I'm the founder and CEO of NAB Robotics. We build on-demand self-driving e-bikes and golf carts for cities, so really focused on being the autonomous solution for short-distance transportation. The cool thing about it is because we're focused on these shorter trips and at slower speeds, we don't have to use the same type of hardware and expensive sensors that Waymo, for example, uses.

can offer really affordable rides. So I always tell people really trying to apply Moore's law to the cost of transportation, really trying to pass that cost savings on to riders.

Mat Sherman (01:38.55)
Totally, my first impression of you is I was like in a room with a of people and someone brought you up and I went to your website and I'm like, this is the coolest thing I've ever seen in my life. literally, if you go to their, what's, ride now.co, is that your website?

Danny (01:46.741)
Thanks.

Yeah, ride nav.co.

Mat Sherman (01:55.864)
That's it. It's just literally just like exactly what you just said, but it's presented in such a fun way. I'm just like, this is badass.

Danny (02:03.987)
Yeah, I appreciate that. Yeah, we took a, feel like there are a lot of cool robotics videos out there. So took a lot of inspiration from like, serve robotics, their website and some other ones.

Mat Sherman (02:14.498)
Did you hire a professional videographer to get it done or did you just do it on your iPhone? How did you get the video done?

Danny (02:20.275)
Yeah, so I tried to shoot it all on my iPhone initially. So the robots you see there are actually remotely piloted. So I tried doing it myself and was remotely piloting the robots that just did not come out well at all. So yeah, I ended up hiring a professional videographer to do it and like cut it a certain way.

Mat Sherman (02:37.038)
Nice. There's some things that we can do ourselves, but I think there's some stuff where it's so much better off having a pro do it. I've done three companies and they all essentially are the same type of logo. It's just black and white with just the logo in the middle. I'm not really a brandy. I don't necessarily need a cool logo, but maybe I'd be better off if I got one professionally done. I don't know.

Danny (02:47.893)
the

Danny (02:52.307)
Yep. Yes.

Danny (03:00.597)
But you did those logos yourself, right? were telling me? Yeah. That's him.

Mat Sherman (03:05.474)
Just Canva, dude. I just like light solid, it good on Canva. I think the logos look awesome. And I think if I ever start another company, I'm just gonna do the same thing. So think they look good enough.

Danny (03:13.333)
I think they're good, man. I was really impressed with the logo. But yeah, it's easy now. I mean, you can just run it into an AI model, just text to image generator. So it's pretty easy. yeah, same thing. I designed them all myself. But yeah, probably would be a bit better off if we got a Pro to do it.

Mat Sherman (03:33.752)
Well, so yeah, that makes sense. So how'd you get into this? Like this is pretty cool. I've like beta tested this, which I can go into in a bit, but what got you into wanting to build like short term transportation robotics or like how would you frame it and like how'd you get into it?

Danny (03:42.089)
Yeah.

Danny (03:50.42)
Yeah.

Yeah, yeah, absolutely. Yeah, I frame it in just that way, right? Like, autonomous solution for short distance trips. yeah, basically the way I got into this, so from Orange County, California, so the Newport Beach area. I try to be as car-free as possible, but it's just not realistic to be completely car-free down there, right? mean, cities are so spread out, you need to have a car. know, biking is really important to me, so.

When e-bikes first came out, really kind of changed my life because it expanded the range in which I could travel around the city on a bicycle. A lot of the big e-bike manufacturers are in Orange County too, so they were really accessible. So, ended up buying a ton of different e-bikes. I still had a lot of issues with them. owning a high quality, reliable e-bike that's not going to break down on you like every four weeks is really expensive. They can be thousands of dollars. Still was having to carry it around with me everywhere.

and kind of store it. So that was really burdensome. And like the biggest issue I had was like, you're going to think that like I'm the unluckiest person in the world, but I had three brand new e-bikes stolen within like the span of six months. They were like locked up, had really secure locks on the beach. Like it was, it was insane. So like after that, that last time I was like, okay, like this is kind of a joke. I just want something that's like a lot more convenient where I don't have to worry about theft.

I don't have to worry about kind of bringing these around with me everywhere. I kind of want like an Uber for a bike, right? And yeah, so that's how I kind of got into it. I've also been building robots on the side. It's kind of a hobbyist for a long time and yeah, kind of decided to combine my love of transportation with robotics and do now.

Mat Sherman (05:36.686)
Can you just talk about the ideal end state with the bikes and then talk about with the golf carts? How would it work once it's all done with an autonomous bike? What would that flow like for someone?

Danny (05:52.765)
Yeah, yeah, definitely. you know, our main goal is to basically build like autonomous kind of petty cabs for the bike lane. So I mean, you tap a button, you know, this vehicle for the bike lane drives to you, you get in, you know, can tell it to go wherever you want. And then when you get out, you just end the ride and it will go pick up another another rider to we plan on selling these outright too. So like if you buy an autonomous golf cart, for example, you know, you take it to like

date night or whatever, you go to a restaurant. You can kind of just flip a switch, turn it into a robo taxi, have it start picking people up for you while you're at the restaurant so you can make money kind of while you're out. And then when you need it, you can just kind of hail it back and it'll come back and you can just kind of take it home. So that's kind of the main vision where we see this going.

Mat Sherman (06:45.046)
It's wild, think if you would have said that 10 years ago, I think people would have thought you were like smoking crack, but today, that's like a thing. It's not a thing, but it will be a thing within a few years, like with Tesla and whatnot. But it's not like Tesla is gonna have a monopoly on everyone having a thing that goes out and makes money for them, right? It's like, there's gonna be dozens of brands that do this, but it's still early enough where I feel like there's some interesting...

Danny (06:51.529)
Yeah.

Yes.

Yeah, totally.

Danny (07:05.597)
Yeah. Yeah.

Mat Sherman (07:12.6)
There's like interesting stuff to do in this space. It's pretty exciting.

Danny (07:14.677)
Totally, Yeah, and the cool thing about what we're doing is because these vehicles are so light, like I said, you don't need as much expensive hardware and sensors on them because they are moving a lot slower. I mean, you can make them for a couple thousand dollars really. Whereas, know, Tesla, you're gonna have to buy those robotaxis for, you know, like 10, 20, $30,000, right? They're gonna be pretty expensive. So yeah, it's a very affordable.

way to get this going.

Mat Sherman (07:45.976)
So help me understand, so I live in like North Scott, let's just say for example's sake, I live in Old Town, Scottsdale. So for people that don't know Old Town, which is most of you listening, Old Town is like this fairly walkable part of Phoenix, Arizona. Golf carts are already there, you know, in the evenings and during the days. And it's like, it's kind of a square, this is a gasp square, three quarters of a mile where it's kind of happening. Would I, so I would like you use this out to dinner in Old Town. Would it essentially work just in Old Town or could

that like could also be taken to Tempe, which is another city or taken to Norris Castle. Can you probably walk like if you haven't figured out and also this I mean it's kind of like a little you know it's obviously it's a little bit of vision but it's like could it just like go wherever someone takes it or is it stay in a certain geography?

Danny (08:22.772)
that.

Danny (08:31.699)
Yeah, so it can. So a golf cart, it's limited to roads that are 35 miles per hour and under. But like when we were, when we first got, so like I said, we're from Orange County. We recently moved to Arizona. And when I first got here, I literally scoped out every single road in Scottsdale, in Tempe, in downtown Phoenix as well. And yeah, they're limited to roads that are 35 miles per hour and under, but.

I want to say like 60 to 70 percent of the roads in each of those cities are basically have that speed limit. So you can totally take it from Scottsdale to Tempe or like downtown Phoenix. It's totally feasible.

Mat Sherman (09:13.262)
How'd you, how'd you, why'd you move from LA to Phoenix? How did that come to be?

Danny (09:18.145)
Yeah, yeah, yeah, it's a good question, right? I mean, yeah, not a lot of people move out of like, kind of beach town, Orange County. But yeah, so we were gonna move the company to Arizona anyway. So basically, the self driving laws here are very, very minimal, to be honest. It's really, they're trying to get companies to be more kind of.

technologically innovative here and really trying to get self-driving going here in Arizona. So, you know, to do this in California, it would have been very complex, a lot more regulatory hurdles we would have to go over. So, yeah, kind of plan was to always move to Arizona. And we're also, we also like ASU as well as like kind of a good first launching place. School is one of the biggest college campuses in the US. So getting around for those students is really difficult.

Mat Sherman (10:15.534)
where did you move to when you moved to Phoenix? You get what I'm alluding to. It's very relevant because where do you currently live and how does that impact your life and what you're doing with NAB? Does it inform you of stuff?

Danny (10:17.301)
Yeah, yeah, yeah.

Danny (10:25.299)
Yes.

Danny (10:30.153)
Yes. Yeah, yeah, absolutely. So I'm living at cul-de-sac right now. For those of you that don't know, Ryan Johnson is the founder, CEO of cul-de-sac. Cul-de-sac is basically building car free cities in the US. So it's awesome here. It's kind of like a perfect match for what we're doing, right? And yeah, that's what we're doing. Yeah, basically Ryan is a...

is a very forward thinker and really sees the future of where this is going. Essentially, when autonomous vehicles become more prevalent, there's really not going to be a need for as many parking spots as there are in the US right now. I think it's like five parking spots for every one car in the US. It's freaking insane. That's why just the way people get around cities and live around cities is so crazy.

What Ryan's doing is he's really kind of eliminating the need for a lot of those parking spots and when you do that and autonomous vehicles become more prevalent you can really rethink the way we build and design these cities and we can really make these things more walkable. So that's exactly what Ryan's doing.

Mat Sherman (11:41.806)
It's wild he was like out in 2018 2019 doing interviews for this cul-de-sac company and like he saw this shit happen Maybe it was 2020. I don't know but like it was it was no it was 2018 because it when I had my last company pub loft Yeah, and they did the interviews in this office called the Wayne Smith so funny and lots of history but anyways

Danny (11:47.699)
Yes.

Danny (11:56.307)
Yeah, yeah, I think that's when he did YC, I believe. Yes.

Yeah.

Mat Sherman (12:06.294)
The fact that he saw this coming from that long ago was wild, but what's even more wild, as I was listening to an interview with Shervin Pishavar, I'm gonna screw up his last name, one of the investors in Uber, he was at Menlo as well, a big investor.

Danny (12:08.149)
Mm-hmm.

Mat Sherman (12:22.03)
And he was saying Travis Kalanick's original vision for Uber is to build Uber into a world where no one owns their cars either. I'm just like, Travis, like, that's wild. Like, people have had these visions for a decade, but it seems like it's finally starting to come to fruition, which is so exciting.

Danny (12:30.74)
Yeah.

Danny (12:40.245)
Totally, totally. Yeah, it's finally becoming more realistic. I think Sam Altman was working on YC cities for a while and they were trying to do something similar to what Ryan's doing as well. Google had Sidewalk Labs. I don't know if they still do it, but they were kind of working on something like that. And then you also have Toyota's Woven City that are kind of building like these futuristic cities with autonomous robots and personal robots.

We're definitely moving towards that way. And cities are becoming on board as well. mean, they're trying to basically plan and design around people, not cars anymore. So that's kind of what they're prioritizing.

Mat Sherman (13:21.144)
so wild, especially with Waymo doing its thing in the Phoenix, not just Phoenix, but we're the biggest market. Hear that San Francisco? We're bigger than you.

Danny (13:24.255)
Yeah.

Danny (13:29.129)
Yeah.

Mat Sherman (13:31.084)
It's just like, this is totally gonna happen. There's gonna be cities that don't cater to cars. They're only gonna cater to people. It's a beautiful thing and that lines things up perfectly for you as well because not everyone wants to spend the money to take a Wemo. Some people enjoy different types of transportation. Just like now there's different types, but just none of it's autonomous. But in five years or one year or less, maybe there's other autonomous options with nav powering some of them.

Danny (13:36.767)
for sure.

Danny (13:52.585)
Totally,

Danny (13:59.125)
Yeah, absolutely. And I think when these vehicles become autonomous, it really allows us to rethink the way we design these vehicles, the experience that we design these vehicles around. mean, you could have certain AVs for different type of experiences. So yeah, it really lets you get super creative and yeah, yeah, could get pretty crazy.

Mat Sherman (14:24.226)
I feel like we've talked about this a little bit off air, just like in our conversations, but I'm curious to hear the way this works, the way you're able to do this with a bike and with a golf cart is that it's really just technology that you put into the, these are very non-technical terms, but you just plug it into the system and then it just can drive itself? It's probably not that simple, but it's probably not that far away from that. Can you walk me through how you're able to even just do this without giving up your secret sauce?

Danny (14:52.393)
Yeah, yeah, absolutely. So I mean you still have to train these models. You still have to map out different roads and cities. But the cool thing about it is, mean we basically built these kind of prototype e-bikes that we have for just like a few thousand dollars with just cameras and a LIDAR. And yeah, I mean it's because you're not dealing with as many edge cases as you are with Waymo for example. I you don't have to have these things be, you

go on the freeways or deal with high traffic situations. You're really not dealing with as many edge cases. So it's really kind of our goal is to make this kind of plug and play where you could put it on e-bike, you can put it on a golf cart, a bit like Comma AI is doing right now, kind of with cars.

Mat Sherman (15:39.694)
I'm not familiar. Remind me what comma is doing

Danny (15:42.461)
Yeah, so George Hotz, have you ever heard of him? He's like a big kind of self-driving guy. He's like always talks shit to Elon Musk on Twitter. He's a pretty funny dude.

Mat Sherman (15:50.52)
Wait is this L-H-O-L-T-Z? That guy? Halt.

Danny (15:55.025)
I believe it's H-O-T-Z. I believe. I could be wrong. Yeah. He's a funny dude. If anybody hasn't heard of him, you gotta check out his Twitter. He's a funny guy. But he's building these little kind of like hardware components that you plug in to a car's CAN bus and you can make it autonomous. It's pretty sweet.

Mat Sherman (15:57.614)
I don't think so. think I'm thinking of someone else, but okay. I gotta find him

Mat Sherman (16:17.416)
Is there a world where you can make anything on wheels autonomous? I wanna like, this is kind of a stupid question, but it just popped up. If I wanna take my Lego set that is a little mini car, I wanna take my little, kid's car, which has wheels and stuff, and can install a motor and make that thing autonomous. I mean.

Why would I do that? I don't know. I'm just really trying to understand the technical limits if there are any of this kind of stuff.

Danny (16:46.409)
Yeah, you can. I think it's totally feasible to do something like that. I mean, you need a computer, a, so you need a computer, cameras, obviously a motor, right? But the cool thing about this is, I mean, we, the first prototype we built, we took a Segway 9Bot go-kart with, again, if anybody hasn't seen those, they are sweet. Like, they go up to 30 miles per hour. They're really sweet go-karts. We took that and we basically made that autonomous. So we just added,

computer and we added four cameras and that's all we did. yeah, essentially if it has a motor computer cameras you can do it. That's why, excuse me, Nvidia is crushing it right now. I mean they're making these tiny computers that are like $500 but they're just immensely powerful and you can run very complex AI models on them. So it's definitely becoming feasible.

Mat Sherman (17:40.654)
Well, what did they they just announced like a $3,000 computer? Did you see this like what do you know? Look as people know on this podcast quite obviously like I'm not a technical person so I asked like, you know these basic questions to technical people and people answer them and that makes me smarter but like

Danny (17:44.861)
I did, yeah.

Mat Sherman (17:57.622)
So I know Nvidia is primarily known for a lot of things, but the GPU I think is a big one, right? Help me understand in super layman's terms, these little computers they're making, they're obviously not competing with Apple, unless they are, but help me understand why are they putting out these $500 computers? What can you do with them? What about this $3,000 computer? What can you do with that? Can you kind of help me understand what they're doing here, at least to your knowledge?

Danny (18:03.731)
Yeah.

Danny (18:22.485)
Totally yeah, so I mean the main thing that they're doing is they're increasing the compute for those computers So you know what like a Waymo car it has several different computers like that they obviously build their own But they have several different ones in you really have to take the camera data that you're getting And the lidar data and every other all the other sensors and kind of fuse that And you need multiple computers to do that so when you increase compute

You can get that going on one computer and really reduce the hardware costs on your vehicle or on your robot.

Mat Sherman (18:58.156)
So the $500 computer, in itself increases compute, because there's a compute, for lack of a better word, like on that computer. Or is that computer just able to store more compute, which means there's more power that can be pushed to wherever that computer is plugged into. Again, all nontactical terms. Yeah. That make sense? And then it's the $3,000 one, just like that, like just can store that much more compute. It's just like more compute, essentially.

Danny (19:16.814)
Yeah, yeah, it's able to store more compute essentially. Yeah, so that's the gist.

Danny (19:27.369)
Yeah, I don't know the exact amount of compute, I want to say it's like two to three times more.

Mat Sherman (19:34.926)
Do you, I feel like I need to do a series of podcasts where I just ask AI people these questions, I understand more than the average person at AI, when you say compute, it's pretty much like power, it's like all this is running on compute, but I'd like to just understand more. How do you learn about, how do you know these things? I guess you're technical, so you just learn online or stuff. How have you educated yourself on all this stuff?

Danny (19:42.026)
Yes.

Danny (20:04.137)
Yeah, for me, so yeah, my background's more on like software, so I didn't have a ton of experience with the hardware side. yeah, I think with any engineer or any entrepreneur, any curious person, mean, with me, I just learned everything online. Like watching videos, speaking with people, speaking with my friends, and all my friends, that's how they learn. I have a lot of friends that...

have gotten PhDs in robotics or computer science or even like postdocs and like they didn't learn anything like anything like thing relating to building product in academia that they really just had to kind of teach themselves and just start building stuff so yeah anything yeah I see that I don't really understand I'll just look it up and start talking to people watching videos

Mat Sherman (20:54.55)
It's wild, I've started using V0 for coding and I'm not, I'm like not.

Danny (20:58.165)
Yeah.

Mat Sherman (21:00.174)
100 % there yet. like there's some stuff I need to do in the terminal But like the thing is I'm starting to google, you know different git commands and I know I have a github and now I have all these things and it's like I think cuz I feel like chat GPT lowers the barrier to learn things cuz all you gotta do is be like here's like here's a screenshot of an error What's what am I doing wrong? And just it just tells you it's like it's quite wild like the cow the knowledge just you know, the knowledge It's like knowledge sharing that's gonna happen. That's already happened that will continue to happen that it leads me to wonder

Danny (21:08.297)
Yeah. Yeah.

Danny (21:18.761)
Totally.

Mat Sherman (21:31.508)
like where's the scarcity in like, or scarcity in like, you know, five years and I think it's just like relationships, but you know, you never know, I don't know.

Danny (21:37.971)
Yeah, it's insane. mean, it's gonna be, the next couple years are gonna be crazy. But yeah, mean, chat GPT, whenever I run into an issue, I'll just ask it for help with programming. It's a really quick solution. think what I used to do is having to post a question on GitHub or post a question on some forum. had to wait like three days for somebody to get back to me. It was a freaking, yeah, exactly. Yeah, it was a freaking nightmare. So yeah, I'm glad that these platforms and these products exist.

Mat Sherman (21:58.904)
Yep, stack overflow rate.

Mat Sherman (22:08.568)
So let's go back to where you're at and like why someone listening up there like, this is pretty cool. I want my own bike. I want my own golf cart. Do you have as of today, well, just two questions. Six months ago, how long do you think it was going to cost you, how much do think it was going to cost you to put these things out? And has that cost come down over time as like the more, you know, all the companies have been building and trying to get the cost to compute lower?

Danny (22:36.073)
Yeah, yeah, in terms of like the individual vehicle or like at scale. Yeah.

Mat Sherman (22:38.894)
Yeah, I mean, at some point I want to buy one of these and you have it roll around an old town and take out all the drivers that charge $50 for half a mile. I want to be like, oh, we'll do it for $10. So I just got to be able you so it's almost like for people that are curious about that standpoint, how do you think about cost? And do you kind of wait for a certain technology to come out so you can lower the price? like, I don't know, just how do you think about that?

Danny (22:44.735)
Yeah.

Danny (22:51.669)
Totally. Yeah, totally. Yes.

Yeah.

Danny (23:07.071)
Totally, yeah, yeah, it's a good question. So like the golf carts, for example, like when we first started looking at it, like I'd say like six months ago, I mean, we thought it was gonna cost around like $40,000. Like I think you and I had like a conversation about that a while ago. I mean, that's crazy expensive as a first prototype. mean, maybe that's okay, but like obviously not gonna work at scale. So yeah, like.

Continue to kind of work on it. The biggest issue is you kind of have to make a custom drive-by-wire system Which essentially means that everything is controlled by ones and zeros There's no mechanical system that controls the acceleration and braking and steering, right? So yeah, obviously have to you know build that and yeah get that working in a safe manner

But yeah, I was just talking with a manufacturer that does a drive-by-wire system a couple weeks ago and we can build one of these golf carts for probably $10,000 right now. And yeah, it's a bit more expensive than like a new golf cart without any self-driving technology.

Mat Sherman (24:20.878)
just thinking like 10 years from now, like all of old town dominated by robots. But that's the future, man. Humans are flawed, right? Robots aren't as long as they're trained on the right data.

Danny (24:24.969)
You

Yeah.

Danny (24:33.909)
Dude, yeah, dude, it's insane, man. mean, the official statistic is there are 100 car deaths every single day, and most of them are result of human error. like, mean, it's like driving. You think about every time you drive, right? Like, how many people do you see like on the phone, like talking or like on the phone texting? Like, it's insane. I was at the Arizona head of transportation department's office.

a couple days ago and yeah we were just talking about saying the same thing. mean it's, she kind of can't wait for the day when like no one drives in.

Mat Sherman (25:10.734)
Is there gonna be a time where it's gonna be illegal to drive or you need a different type of license to drive because most of the cars out there will be operated autonomously?

Danny (25:18.762)
Yeah.

Yeah, that's a really good question. So I think what's going to happen is that insurance will just be so high that in order to drive a car, it'll be more of like a hobby thing for really wealthy people that still kind of get their fix on driving like really high-end sports cars and things like that. But yeah, think it's definitely the cost to actually drive a car is going to be really high.

Mat Sherman (25:44.608)
And why would, what would drive up insurance if humans are flawed and robots aren't? like the chances of a human screwing up a robot driven car goes up or I guess help me understand why would insurance rates go up over time?

Danny (25:56.029)
Yeah, yeah, that's exactly right. Yeah, so just basically because it's so unsafe for humans to drive and there's so many accidents as a result of human error actually that them driving these cars that, yeah, to ensure them it's going to be very expensive. I think you compare that to how many accidents Waymo has been in the Phoenix area over the last, what, like two, three years. I think it's around like seven or eight and most of them have been the fault of other people.

Mat Sherman (26:26.542)
Yeah, it's funny. It's actually not funny, but I saw a video on Twitter or something of like a Waymo. No, it even autonomous related, but I'm still gonna share the story. The video was like, how to deal with insurance scammers or whatever, and someone's driving, and it's like in the middle of the night or whatever, then someone just added it where it runs in front of the car, and the car hits the person. Clearly they were trying to get hit. Then someone magically on the other side was there to pick them up and be like, oh, are you okay? And the car just drove off. It's just like, I'm not dealing with this.

Danny (26:27.957)
Yeah.

Danny (26:46.783)
Yes.

Danny (26:52.544)
Yeah, for sure. Totally. Yeah, yeah, it's hilarious. Yeah, I remember, like the first time I got into Waymo, I was definitely trying to mess with it. I was trying to grab the steering wheel, like trying to just do crazy stuff. Cause I mean, I'm just like so impressed by the technology. mean, it's really, it's literally like magic. But yeah, I see like people trying to mess with it all the time. It's fun.

Mat Sherman (26:56.448)
It's pretty funny. It's pretty funny.

Mat Sherman (27:19.566)
Yeah, I know someone that tried that like honked actually actually I'll leave them anonymous because there's someone that I'm to but They honked on the horn because like they wanted to honk on the horn there like there was a car that was and then I think the way vote kicked him out like they just like it just stopped and Pulled over and it was like you got to leave you honked the horn. I was like, oh shit

Danny (27:24.265)
No.

Danny (27:30.26)
Yeah.

Danny (27:37.641)
Yeah, that's awesome. Yeah, it's hilarious. Yeah, it's always fun. I know initially they were testing with remote pilots, so yeah, I'm sure they could see everything that was going on in the car, so yeah, it's pretty

Mat Sherman (27:51.074)
for sure. Is there anything else you want to share in relation to Nav that is like something that you maybe learned as you've been building it or something that you think is like interesting about the trend in the market or anything like that?

Danny (28:02.389)
I mean I can talk about the go to market stuff that we did if that would be helpful for your listeners.

Mat Sherman (28:10.03)
Yeah, well this is how we met in person for the first... Yeah, it's how we met in person. Yeah, please share.

Danny (28:15.049)
Yeah, believe so. Yeah, yeah, you were one of our first beta testers. So yeah, we did a beta at ASU a couple months ago and that wasn't our first beta. So we actually did a beta at UC Irvine in California, which is the closest college campus to where we're from. And it was horrible, man. Like it was really bad. So we actually tested it with the robots. The robots, again, we tested these with prototypes, so the reliability was pretty low.

So we took a lot of those learnings and kind of applied it to the ASU beta. And what I think, so just to give everybody context, at the ASU beta, we did that in, I believe, like mid-August when classes were first starting. And I actually bought a van and bought a bunch of different e-bikes and manually delivered them to students that requested rides. So we didn't even use our app either. We just used straight text messaging.

Mat Sherman (29:11.118)
So I got, so I did this and it was pretty like professional and so I text this number to you know to book and I text the number and I'm like wait this number is saved in my iPhone. This number is just straight up Danny's number. It was your number which was hilarious because that's exactly what else would it be?

Danny (29:21.365)
Yeah, yeah, yeah, I remember. I remember you were super surprised. You're like texting, you're like, hey, pick me up. And like, wait a second, this is Danny's number. Yeah, it was hilarious. But yeah.

Mat Sherman (29:35.662)
because I felt kind of rude because I'm just like well I don't know like it's just this number maybe it was automated I don't know but it's just it's very founder it's scrappy level 10 the fact that it was to you which is great

Danny (29:45.491)
Yeah, yeah.

Danny (29:50.281)
Yeah, super scrappy. Yeah, I didn't want to buy a phone or set up some remote wireless system to route the calls or the text messages. So yeah, it was super scrappy. So basically what we learned is kind from that first UC Irvine beta is that the most important part of a minimal viable product is kind of the viability of it. Like how viable is it? Because we couldn't even really use

these things at UC Irvine when we were first testing. So that's kind of what we learned at ASU is look, like what do we know is gonna work? Yeah, buying a bunch of different e-bikes and just kind of replicating the on-demand experience that we're providing. We also like hired like 15 interns. I think you met a few of them, but those dudes were just straight up killers. mean, they knew that they were part of a bunch of different clubs and they were responsible for a lot of the rides we gave them.

referral kind of payouts where every ride that they referred they would get a portion of the revenue from each ride. yeah, they end up making a couple hundred dollars throughout the kind of four weeks that we were there. But yeah, the viability is kind of the most important thing that we learn. With software it's a bit easier, but with hardware and robotics it's tough. So that's how we kind of just kind of manually did it at ASU.

Mat Sherman (31:13.326)
I think the fun thing about the beta test at ASU was, you know, my wife and I used this, and by the way, the bike you gave me was badass. It wasn't 30 miles per hour, it was like a mini motorcycle, but it was definitely an e-bike. But it was funny, because we got to our location and we just like left them there. it was beautiful, because in theory, and in practice, once it's ready, they'll just drive off, right? But it's just a really nice user experience, to be honest, and not having to worry about, do I lock up my bike? Is someone gonna steal it?

Danny (31:23.12)
Yeah. Yeah, it was sweet.

Yes.

Danny (31:34.089)
Yes, exactly.

Danny (31:41.033)
Yeah.

Mat Sherman (31:43.34)
If someone tries to seal it and it gets damaged, know, it's just like ride and leave pretty pretty easy

Danny (31:45.801)
Yeah. Totally, totally. Yeah. Absolutely. Yeah. Like the other alternative, like the other obvious alternative is the scooters, right? It's just like the biggest thing with the biggest complaint that people have is, I mean, you have to leave those in certain areas now. They're highly regulated by city. So the cool thing about what we're doing is you can literally leave the bike anywhere because it's going to immediately drive off, pick up another rider. It's never idle.

in public spaces like the scooters and because of that we're not regulated the same way either. you know cities make these scooter companies pay very high daily fees and annual use fees so we don't have to do any of that. We kind of have free free reign to operate where we want, when we want, how we want because these things are autonomous.

Mat Sherman (32:33.71)
Do you have to go through the cities to enter a new city? Like do you have to talk to city of Tempe to, or someone in city of Phoenix to do activities here?

Danny (32:43.069)
Yeah, yeah. So every city is different, but we definitely go and talk to each city first. So we like to have a regulatory first approach. So I think what the scooter companies did and like what Travis Kalanick did at Uber, like they just kind of use brute force. I think that worked for Uber, but I think cities really smartened up after that and they weren't going to let tech companies come in and just kind of change the way people kind of move around the city. So that's why you saw a lot of over-regulation with the scooters.

We kind of took their learnings and really like to have a regulatory first approach and speak with them, get them on board, make sure they're okay with everything, kind of tell them exactly what we're doing and yeah, kind of do it that way.

Mat Sherman (33:26.254)
Yeah, that makes sense. When I think of Bird, I just think of David Sacks. That was his first investment out of Kraft Ventures. Woo! And like, it didn't end up doing, I mean, it went public, it didn't end up doing quite well. Do you have like an opinion on like, did overregulation kill those companies?

Danny (33:32.082)
Okay.

Danny (33:35.966)
Yes.

Danny (33:43.891)
Yeah, for sure. think if you speak with any anybody who was part of those early teams, yeah, for sure. mean, like they like people don't understand how much they have to pay in fees. I mean, like they have every time somebody uses a scooter, like part a portion of that revenue goes to the government. They also have to pay. It kills it. Yeah, they have to pay annual fees as well. They're capped on how many scooters can operate in certain markets, too. So

Mat Sherman (34:03.276)
It just kills a company. Kills it.

Danny (34:13.749)
Yeah, they totally I mean that like you saw like everyone saw how strong the product market fit was But you know for those 2018 when they first started the prices were low and they still could make the unit economics work But because they're just highly regulated they have to these high fees either raise the prices a lot They don't have scooters in in a lot of the popular areas. So Yeah, it just kind of killed killed what they were doing

Mat Sherman (34:41.226)
Makes sense and I guess it means they're too early because I feel like there's more cultural shifts that are going towards a big a post-car society now mainly I mean is is that just like single-handedly Tesla and Waymo? No, because I think that Rian Johnson knew about this I mean it started with uber I guess you could say and they couldn't do it but the torch was passed to these other companies that have kind of carried the torch to like to the to these finish line of autonomous driving you know which is like Waymo and Tesla and whatnot

Danny (34:47.743)
Yeah.

Danny (35:08.735)
Totally, yeah, totally. think, like even like, you know, I would rather take a Waymo than take a public scooter that's kind of out on the street now. So yeah, I think they totally, totally ate into the market.

Mat Sherman (35:18.83)
I've been hit by multiple cars on scooters. It's horrible Like the situation when the car is turning right, but they don't put on their blinker and I'm like going on a sidewalk I Fall into the ground. I'm a big guy like it was not I was like not not optimal and then he's like yo Are you okay? And I see his car is like kind of like messed up. I'm like, yeah, I'm fine. Good luck

Danny (35:22.665)
Are you serious? Yeah.

Danny (35:28.285)
Yeah. Yeah.

Thank

Danny (35:39.157)
Yeah, I know man. It's crazy like biking is super dangerous and like scooters and dude when we were doing our beta I was delivering a bike to somebody I was riding it to them and a car freaking hit me like from behind same thing I was turning right and they just kept going and Yeah, I tried to be a savage and rally and just kind of continue driving it to the rider But I was pretty pretty hurt after that

Mat Sherman (36:03.982)
does, I mean, how do you think about that for NAV? is that, you know, people will get hit, like that's, I mean, or not, in a post-human car society, I don't know, like how do you think about these risks?

Danny (36:14.037)
Yeah, yeah, so that's the cool thing. mean, these things have cameras in them all around the bike. So like they can see what's going on, know, 50, 100 feet, 200 feet around the bike. like we plan on having a little UI on the bike that actually shows everything going around.

I'm kind like a Tesla UI.

Mat Sherman (36:46.306)
Yeah, that makes sense. And then I guess like...

If you had to pick, and maybe it's not a fair question, and you can say that, but like, do you wanna be bike focused? Do you wanna be, keep forgetting the name, do you wanna be a golf cart focused? Do you wanna be multimodal focused, so you multiple types of vehicles that you run? How do you think about that? And then we'll kind of round it out with a few more questions and call it a day.

Danny (36:52.501)
Thank

Danny (37:03.925)
Thank

Danny (37:09.631)
Yeah.

Danny (37:15.293)
Yeah, yeah, for sure. Yeah, ideally we'd love to have different options, right? I think that's the cool thing about transportation and in what Uber did is they had multiple options, right? You had the high-end black version, which they first started with, and then you had UberX and UberPool and different options. So yeah, I think in an ideal world, you definitely have different options. But from a unit economic standpoint right now where the cost of hardware is, I think we'll probably start with golf carts, essentially. We can make the unit economics work.

there really well. can offer a little bit higher prices than the bikes. And yeah, just a bit easier to build right now. So that's probably start off with the golf carts. But yeah, like I said, it'd be cool to have other options, know, with bikes and petty cabs and things like that as well.

Mat Sherman (38:02.638)
Cool, and I guess what does the next year to two years look like for Nav? I think if everything goes as planned, where are you at in one or two years?

Danny (38:15.261)
Yeah, yeah, so the next year or two, really just making our prototypes more reliable. It's complex to build a prototype, but it's not as complex as making these things more reliable at scale. that'll take another year, most likely. And yeah, really trying to fine tune our self-driving algorithms to make sure it's safe, because that's the other thing, right? You have to make sure these things are extremely safe.

building in redundancies in the models and in the hardware. So yeah, that'll probably take up the next year or so.

Mat Sherman (38:55.054)
Do you, let's say you train a data set on like Old Town Scottsdale and then you plug that into like a golf cart. Let's say you buy, let's say I buy another golf cart. Do I need to train it all again or do I take that same data set and just plug it into the second golf cart? Does that make sense?

Danny (39:12.361)
Yeah, can just, yeah, yeah, exactly, yeah. That's the cool thing about it. You can just take that data set and plug it into the second golf cart. So as long as you have the same camera setup and other sensors set up, it's really just a plug and play thing.

Mat Sherman (39:27.886)
Makes sense. I think this is the coolest idea ever. And it's cool because you're technical. You can actually do it with the right financial partners and the right market conditions. This can be a real thing. And it doesn't even take a leap of faith to get there. And I think that's pretty exciting.

Danny (39:29.225)
Yeah.

Danny (39:36.425)
Yeah. Yeah.

Danny (39:44.277)
Yeah, totally. Yeah, yeah, like we can definitely build it. I think what we'll probably end up doing for that for it, we'll probably just build a golf cart with the self-driving capabilities integrated into it and just try to sell that outright initially and maybe just try to sell a couple of those kind of pull a Michael Dell, what he did when he was getting Dell computers going and it's called dorm room, kind of in the early days. yeah, just, you know, it'd be nice to sell, you know, 10 to 15 of these in Scottsdale. And then when people want to take

turn these into robo taxis, can kind of just flip a switch. They can start picking people up on weekends and they can start making some money.

Mat Sherman (40:21.678)
One last question that popped up

just as you said that is would you need other people to like, the other people would need to like have an app to call these golf carts. So let's say I buy a golf cart and I'm out to dinner in Old Town and my golf cart is now free to help other people. Do people need to book it on an app? Could they like flag it down? guess, I don't think I've ever asked you this actually, just like in general. How can someone like get, like how does someone book a ride on a golf cart? At least as of now.

Danny (40:32.853)
Mm-hmm.

Danny (40:43.401)
Yes.

Danny (40:50.697)
Yeah, yeah, yeah, so we have an app. So yeah, you basically just tap a button and it comes to you. same exact thing as a Waymo or an Uber.

Mat Sherman (41:00.482)
Will you allow for, this probably isn't possible, but would you allow for like a similar way that people can hail taxis, people can hail the golf carts?

Danny (41:07.689)
Yeah, dude, think about, yeah, I think about that a lot. I think that'd be a really cool experience. Yeah, kind of bringing it back to the way everything started. you, from a technical standpoint, it's totally feasible, right? Like you train these models to actually recognize when people are kind of like hailing. You could totally do it. I think it'd be a cool, cool feature. Yeah, interested to see what other people think, but yeah, I think it'd be cool and it'd be totally feasible.

Mat Sherman (41:34.638)
feel like you just need to like, I'm out of so many little issues with the little things that would be hard to execute that on, but you need to know who they are, right? So feel like once they hail it, you'd be like, great, we can't move until you download the app, something like that, I don't know. But then that's like added stuff friction, I don't know.

Danny (41:40.147)
Yeah.

Danny (41:46.825)
Yeah, something like that. Yeah, we like the texting kind of thing, like texting experience. Yeah, so like there was a company called Magic. Have you ever heard of them? They came out like, I want to say like eight years ago.

Mat Sherman (41:54.22)
Yeah, totally.

Mat Sherman (42:00.814)
Magic, the authentication company.

Danny (42:05.621)
No, so they did basically like you could text the number and You could get anything so like if you wanted like a sandwich you could just text this number like hey like pick me up You know like a Bay City sandwich, and it would it would like Basically like order it through door-dash and send it to you you you're paying a premium for that, but It was like the coolest experience ever so

I think they're still around. think the founder's name is Mike Chen if anybody wants to look it up. But I think they're still around. They might be doing something a little bit different now. yeah, texting, that's, I love that. I think it's like super seamless. It's fun. That would definitely be a possibility.

Mat Sherman (42:46.016)
Air Garage had a nice flow for texting, like where you pull up to a parking spot, there's a QR code, like you, I don't know, I feel like they allowed you to book a spot without having to download an app.

Danny (42:58.343)
Okay.

Mat Sherman (42:58.424)
which I think I all need to, I mean, I know where they've, you know, I know where their spots are. Maybe I'll go back and try it out. Cause I think that's, I think if you make allowing someone to like use this that hasn't already downloaded the app and like give them a good experience, them having them download it makes, I think makes sense versus, another app to download. shucks, you know.

Danny (43:03.38)
Yeah.

Danny (43:16.265)
Yeah, Yeah, I totally agree. Even like having a QR code that they could scan and maybe get like a payment link. the other interesting thing about that is that, I mean, if you have an app and you're taking payments through it, Apple's taking like, I want to say it's like 40 % right now. It might be a little bit more. Somebody needs to fact check me on that. But it's ridiculous. Android is a little bit less, but it's still very high as well.

Mat Sherman (43:25.965)
right.

Danny (43:45.577)
Yeah, that's why I really like the kind of texting or just like sending them a paymailing.

Mat Sherman (43:50.134)
Yeah, sounds like web app it is. Cool. Well, where can someone learn more about Nav online? Where can someone find you online? You have social media. What's the website? How can someone learn more?

Danny (43:53.532)
we appreciate that.

Danny (43:59.241)
Yeah, totally. Yeah, so website is ride nav.co. So R-I-D-E-N-A-V.co. And yeah, I mean, I'm on Twitter. I don't really use it that much. I just kind of look at like what you want us to say and other tech people that I like. But yeah, I believe I'm underscore Miniatillo on Twitter.

Mat Sherman (44:23.886)
All right, thanks for coming on the podcast.

Danny (44:25.799)
So yeah, thanks for having me. I really enjoyed it.