Ep. 7: Jason Richards

Robotic Food Delivery, Crowdfunding, and Lawmakers

November 16, 2021 · 1:44:13

In this episode, Audrow Nash speaks to Jason Richards, CEO at DaxBot. DaxBot makes a charismatic robot for food delivery. Jason speaks on their delivery robot, their crowdfunding campaign, DaxBot’s revenue model, working with lawmakers to have robots on the sidewalks, and Jason teases their realtime operating system, DaxOS.


Links

Comments

Transcript

The transcript is for informational purposes and is not guaranteed to be correct.


Audrow Nash (A.N.) (0:03)

This is a conversation with Jason Richards. Jason is the CEO of Daxbot, which makes a charismatic robot for food delivery. We talk about deck spots delivery robot, their crowdfunding campaign, that spot revenue model, how Jason works with lawmakers to have robots on the sidewalks. And Jason teases their real time operating system. DaxOS. This is the Sense Think Act podcast. And thank you to our founding sponsor, Open Robotics. Now, here's my conversation with Jason Richards. Would you introduce yourself?

Jason Richards (J.R.) (0:42)

Sure, you bet. My name is Jason Richards. And I am the CEO of Daxbot.

A.N. (0:49)

Would you tell me a bit about Daxbot?

J.R. (0:51)

Sure, yeah. So Daxbot is a robotics company that started out at a research and development lab back 2015, the founder, started a couple other companies, pretty successful guy, and decided that he wanted to create a research and development lab to, you know, what is the next wave of technology that we can get in front of, you know, and so, in 2015, one of the thoughts was, okay, well, we need robots to be able to do things. And there's plenty of plenty of robots in the, you know, in the workforce, you know, more in factories and things like that. But how do we bridge the gap into the next generation of robotics, where you have robots that are on the streets, similar to, you know, iRobot? Before the robots go bad kind of thing? It's like the movie being able to say, yeah, like,

A.N. (1:40)

not the company backs the back. Yeah.

J.R. (1:42)

Not No, no, not that company. No, yeah, the movie, right. So you can tell a robot, Hey, pick up that, and go watch that window. And robots are smart enough to go and do it in human space. So that was the idea. And looking at what things could be most helpful out of the chute, is the delivery seemed like a natural first thing that a robot could do that would be pretty simple, you know, simple tasks that would be helpful to humans. So so that's where the idea came from. And so as of recently, we just got through restructuring companies now we're dexpot, Inc. and now we're able to go after and commercialize and, and have robots now being deployed to different states around the country where you before? Yeah, so before is Nova dynamics. So Nova dynamics was the research and development incubator. And they've got several projects, Dax as the robot was kind of the big one. And so with that platform, we knew that at some point, it has to move out of the r&d Lab and into its own entity that can be commercialized and become something Gotcha. So. So that was the next step. So as of last year, when you know, all the research and development last year, when COVID hit, it was a great opportunity to start getting robots out in the field, in our local hometown, because restaurants, people aren't coming to, you know, restaurants are closing their doors. But people aren't wanting to see people. So it seemed like a great space. And it was, you know, for that beta test, being able to take robots that we have go to the restaurants, pick up food, and then ticket to people's houses, you know, all without seeing a human was really highlighted the need for this type of technology. And it is well received. And it was a great, great beta test for us. So January of this year realizing okay, we've got a product that is ready to go awesome. So commercialized and

A.N. (3:34)

here we go. Hell yeah. Tell me can you describe what Dex looks like?

J.R. (3:40)

Yeah, so Dax is a robot? It's about right. Yeah, yeah, if you're, if you're watching the video, you can see the one behind and it's about three foot tall, and white and black, kind of a reverse tuxedo type situation. And so you see, you see, Dax has got a fully articulated neck so that he's able to look all around and look up and down also has LED panel eyes. So the eyes are able to be expressive, you know, whether it is being able to be hearts or you know, I love you going to sleep blinking, so that it has some sort of emotional response from another human. So that is and then the, the cargo that's the thing that everybody asks, well, if it's the delivery robot, where's the cargo bay, and it's actually inside the stomach. So what you think of the stomach of this little creature, for people that haven't seen DAX before, I'd encourage you to go, you could check out DAX spot.com. You can see pictures of him and that kind of thing, but, but inside the drawer, which is where his stomach is, is where the payload goes. And then inside that is, you know, heating elements. We've got to design for a refrigerator so that we can do that's cool pharmaceutical delivery. Yeah, things that need it, which is pretty. That was a pretty astounding thing from the engineers perspective to be able to build a refrigerator Inside the robot, obviously, it takes up a lot of space with a compressor and, and heat exchange, especially in that type of area. But fascinating, and we did a lot of cool testing with that one. But so that is is Dax.

A.N. (5:14)

And so, to me, Dax, I was gonna say refrigerator, like a little like college dorm freezer kind of thing. Right? It's kind of looks like one of those. And then it has this articulated kind of like hexagonal head that has the LED lights on it. And that head is on a neck that seems to have a few degrees of freedom, which I'm sure we'll talk more about that later. And we can't see it in the video now because of how the cameras positioned. But DAX is on tank treads, right? Yep, gotcha. Correct. And to me, Dax looks a lot like Wally without arms. And like, you say, like, the belly is where you store the payload. So whatever you're delivering, like in Wally Wally was designed to have that trash compactor there and its belly. So it seems kind of similar to me. Right. And the tank treads. Yeah, these kinds of things. Yeah, we,

J.R. (6:10)

we joke about the fact that Dax kind of looks like Wally and Eve had a baby. And that baby.

A.N. (6:16)

That is right, because of the coloring. Yeah. And the sleek design color in the scheme. Yep. Yeah, yeah, exactly.

Okay. Tell me a bit about the neck. Or the head, I guess. Yeah. Yeah.

J.R. (6:28)

So the getting the biggest piece that we wanted to solve was, you know, the human robot interaction piece. There other than delivery robots. There's a lot of people that will ask about that, you know, why? Essentially another delivery robot. And it's it's kind of funny because, you know, Dax was built people are familiar with starship, Dax and starship started almost the exact same time not familiar with starship as far as what a starship? Oh, yeah. So start. Yeah, so starship is a delivery robot on some college campuses. And if you look up delivery robots, if you just Google it, you'll find you know, kiwi, bot, and starship. And there's some others. And essentially the same, the design is exactly the same. It's about it looks essentially like a cooler with a couple of wheels on it. And so it has a payload and it's, it's useful in some applications. But with the the neck and the eyes, it was more of, you know, how do you create something that is going to be accepted in human pedestrian spaces. So so that's where the idea of, you know, articulating neck head, not putting sensors all over the place, and the founder, you know, dexpot really wanted to make sure that it wasn't creepy. You know, there's, there's plenty of really cool, like, mechanically amazing feats that people are coming up with, you know, as far as like mechanics of robots, you know, bipeds, those types of things. But if if, for example, you just saw, you know, two legs, two robotic legs running you down the street, you'd think that Armageddon was happening, I mean, that that would be a creepy thing. So wanting to make something that would be accepted in human space. So there's a lot of work that went into that, of how do we make a robot that people want to have around? So So delivery robots, most of them have that design of a cooler on wheels, and about two feet off the ground. And it becomes an annoyance for a lot of people, because it's it's in the way, similarly, people don't really, people don't even want bicycles on sidewalks, right? And it's a human on that bicycle, but it's annoying and pedestrian space. And so they make bicycles

A.N. (8:35)

that cars have. And so I don't know if it's quite the same thing. I get that they're also annoying, but perhaps it's like, they go fast. And there's the potential to like, get hit by a bike bicyclist or something like this or them for to come out of nowhere. Yeah. Whereas I don't know that small tub robot would be quite the same thing. Or what do you think?

J.R. (8:54)

No. Yeah, the biggest piece had to do with with visibility is not being you know, not being able to see it. If it's it knee height, it's different than if it's it, you know, waist height, or, you know, shoulder catches.

A.N. (9:05)

So it's like a mobile piece of furniture to trip over if it's short, and that could irritating. Okay.

J.R. (9:12)

Yeah. Yeah. And same way. I mean, you know, people don't like skateboarders on on sidewalk. Yeah. Again, moving fast, though. Remote Control often, yeah, some of it could be moving fast. And it's in a way it's becomes a tripping hazard, that kind of thing. And so the interesting thing with Dax is that with all the testing we've done on the HRIS side of it, is you know, especially kids love Dax, it's about the same height and so they can give it a hug, and it's looking them like almost either with very cute and with and with the neck itself. And the reason that we didn't want to creepy with sensors all over the place, people ask, well, it has 360 degree cameras in it, for example. You know, it's kind of an assumption a lot of engineers have, it's like, No, there's not 360 degree cameras in it. And people always want to know why it's like, well, you don't have 360 degree cameras and you know, you can kind of hear that their stuff behind you. And so you can, you know, you have some sense and awareness of what's around you. But you're not a spider, right? You don't have this spidey sense where you can you know that something's behind you. So in the same way, that's where the articulating that came from, you know, you could have something that that knows everything around it in a 360 degree view, or you have something that actually is more like, you know, more like a dog or a human, where you actually have to turn your neck around to see what is actually going on. But yeah, and so even those those pieces, it gives it a feeling that it is actually alive, instead of just being a machine, which why not? It actually is machine.

A.N. (10:42)

So I mean, I'm thinking of this thing, where it's like airplanes don't flap their wings kind of thing. So I mean, we can draw some inspiration from biology. But where do we end Drawing inspiration? From biology? Absolutely. And I wonder, I would think it would be possible to use cleverly hidden sensors and to have the articulating neck, but have the articulating neck not necessarily tied to a vision system. Why Why limit it to having the vision system on the articulating neck instead of just having hidden sensors and this kind of thing?

J.R. (11:22)

Yeah, that's a good question. I don't know, the, the founder when he went through all of this, there was a lot of there was a lot of, you know, reasoning and things like that. And again, making it more a phrase that we use around here is anamorphic. You know, it's not anamorphic. It's not an ad. Cannon morphic. I don't know who doesn't like thrupp amorphic are more like a like a dog. Yeah. Yeah. Yeah, it seems more like it's a pet than it is a machine.

A.N. (11:51)

So anamorphic. So it's like anthropomorphic animal. So when we assume but with cane, but for things like dogs, or whatever it might be, okay, so it's a creature and has some sort of intent when it moves around the world or whatever it might be. Right.

J.R. (12:06)

Yeah, exactly. And so one of the things, for example, is, I believe one of the stories I heard is that if Dax you know, robot goes into a situation and needs to go backwards. You know, we as we as humans, don't walk backwards, right? We don't just like walk backwards into into us, yes, we will turn and say, Oh, here's where we're going to go. And then, you know, if we need to have some field division in front of us, we'll be looking back and forth. So in the same way, wanting to train, train the robot, right train the neural net, so that it's not just doing things automatically. So it actually is showing intent, that oh, here's where I'm going. One of the things that we've heard, we actually, we live close to a university, Oregon State University. And one of in starship, one of those competitors is up there. And I haven't heard, I haven't talked to anybody around that really likes them on campus. And the reason for that is because it has to do with intent. So the device will drive up to a street corner, and then stop. And then all the other traffic, the cars are like, What are you doing? Yeah, what do I do? Yeah, is it gonna cross? Am I gonna cross there's no clarity? No, there's no law about right of way. Yeah, exactly. So there's sensors, no signaling, there's, there's no idea, you know, are you going to go am I going to go and so it's, you know, some of these devices have been hit one guy hit by a train that's been hit by a truck, one's one into a train. And so it's, you know, and it has has plenty of sensors onboard. You know, that's, you know, with that one specifically, it's, that's not the problem. But because people don't know what it's gonna do, they don't have any, any idea versus Dax will get to the street corner, and like a human, it will look at you, and it will watch you and follow your vehicle as it drives past. So you know, oh, this thing sits somewhere. And it's aware of what's going on. And it cannot its head, like, yes, you can go Go ahead, or, you know, shake its head so that people know, Oh, you want me to go? Versus you're gonna go right. And so there was a lot of there was a lot of thought and intentionality into the neck. Yeah, isn't just necessary for the neck sake, but it's more for communication, it looks one of the design,

A.N. (14:09)

it looks to me. So thinking of like the founders perspective, and this kind of thing, perhaps it had something to do with making it so that so I mean, if you have sensors everywhere, and you can get a really good rendering of the world and this kind of thing. You can design behavior that uses that understanding of the world. But in this case, you're kind of constraining the behavior of the robot to be something that seems more natural. In like, it has to go and look behind it. There's something if it's going to go and back up or if it wants to, like look both ways for the street, you have to do that with the head. And it almost so it basically constrains the developers to like, I don't know program more lifelike behavior in it so that it operates in the world. I could see that being a clever way of doing that kind of thing.

J.R. (15:04)

And for for engineers, specifically, mechanical engineers, they would much prefer to have the device to be a cooler, right? Just right. Yeah, that's it, that's a much simpler thing than a fully articulating neck was dual pitch motors. Right?

A.N. (15:17)

Yeah. Or just a sensor information would be probably more valuable. Absolutely. The who cares if it looks like a cooler, but I get your meaning that for like, sensors everywhere and full 3d map of the whole environment that it's in and this kind of thing?

J.R. (15:30)

Yeah. Yeah, absolutely. And we, you know, obviously, we got, you know, that technology to be able to do that, so that Dax is able to learn in 3d space, and you know, instead but, again, more similar to, you know, a human if I'm walking down the street, and I've been down that street before, I have an idea of what to expect, right? I know, there's a curb there. I know. There's a telephone pole. I see. Oh, there's a box there. I've never seen that box before. Okay. New information might not be there tomorrow. Right? So it's not something that that really comes in. But because everything is changing in such a rapid space, again, on a factory floor,

A.N. (16:05)

you mean in the environment? Like walking around?

J.R. (16:08)

Right, yeah. And environment or walking? Right. If you're walking downtown, you know, in a metropolitan area has a lot of moving parts. It's changing all the time. Yeah, exactly. So so being able to have something that is more lifelike to be able to adapt in that environment. And it seems more natural in that environment. Actually, pretty interesting.

A.N. (16:26)

So what sensors are on this robot on the Dax robot?

J.R. (16:31)

Yeah, so despair division is is one of the two it has

A.N. (16:34)

a stereo that has two cameras beneath the kind of chin. Yes, something right below the display. Right. And

J.R. (16:41)

yeah, so yeah, where you see the display of the eyes is where that

A.N. (16:45)

just been, you know, just beneath it. Yeah. Cuz I think I see that a little circles from the cameras that it has. So you have the two cameras. And from that, you can infer some 3d information about what's in front of the robot. Okay. Yeah. Anything else? I mean, I imagine there's encoders and things on the wheels, or the track. Oh, yeah. What, uh, and then all the motors have sensors and things. But what else for perception?

J.R. (17:11)

Is it just those cameras? Maybe microphones? Yeah. Yeah. So it's got, yeah, it's got stereo microphones. See, so you can hear what's going on. Again, kind of that sense. If you're, if you're a human, you can hear somebody coming up behind you, right? You don't have eyes in the back of your head, but you can hear that something is happening. And you're more aware of your environment that way. So, so yeah, those those. It's fun. When I when I talk to people about sensors, a lot of times they're thinking very unilaterally, and they're thinking about, like Lidar and

A.N. (17:42)

things.

J.R. (17:44)

Yeah, exactly. Yeah. Like LiDAR, ultrasonic. I

A.N. (17:46)

mean, I'm honestly surprised. You don't have like a plane or lighter or anything in there. But if you can do everything with vision, that's great.

J.R. (17:53)

Yeah, yeah, it's, it's pretty fascinating, obviously, yeah, the tracks have sensors, things like that, you know, when we're going into, you know, autonomy, things like that. Being able to have tracks to be able to know, even if you lose, you know, GPS signal and that kind of thing, you're able to know, within, you know, centimeter to where you're at in your thing.

A.N. (18:13)

The tracks are very accurate for encoding it. And that's probably because they don't slip as much as wheels might.

J.R. (18:20)

Correct. Yep. Yeah, exactly.

A.N. (18:24)

So Gotcha. Okay, so going back to the neck, you have these two cameras, and those are that's the majority of the bed that is the big perception sensor, the two cameras, and then you have stereo microphones, or whatever it's called, so that you can tell directionality of the sound, right? Yeah. Great. That's that's pretty much it for perception. Other than like, actually, I forget what it's called now, like dead reckoning your position for from the tracks and this kind of thing.

J.R. (18:55)

Yeah. But that's it. Yep. Yeah,

A.N. (18:59)

that's it you've like, are there like capacitive sensors? So it can tell if it's being pat or hugged or anything like this? Or how do you detect like, a person is your buyer interact? I guess the cameras would do that. But do anything for a person touching it or anything?

J.R. (19:16)

No, no, not. We've, you know, talked about that a little bit. It's not, we don't know, how helpful that would be at this point. You know, everything, everything has to have, you know, some lead time and stuff like

A.N. (19:27)

that. What do you mean meantime,

J.R. (19:28)

you know, just which I like to do is that just takes time. Gotcha. Yeah, yeah. Just development time. You know, and so what you know, weighing priorities, which is the most important thing totally, but yeah, with with you know, the vision for example, you know, if if somebody sneaks up behind me and touches me, yeah, I would know that and maybe the robot doesn't at this point might being able to hear this. Yeah, exactly. Somebody is coming up behind me. I can look back and be like, Oh, hey, what's going on? And then you know, we can have a conversation so no, no perceptions. As far as that's gotcha

A.N. (20:00)

is just curious with like interacting with little kids and stuff with like the little kid with patted on the head and would know with being pat on the head. Oh, yeah. But

J.R. (20:09)

people people are hugging it all the time.

A.N. (20:13)

Yeah. Okay, so the neck. Tell me about the neck. So it how many degrees of freedom does it have? What kind of things gonna do? Yeah.

J.R. (20:21)

So not wanting again, not wanting to be super creepy. So we've limited the range.

A.N. (20:27)

So we can't go around like, a hole or something.

J.R. (20:31)

Yeah, we don't want this. It's Halloween. Yeah. Yeah, robot, right. Yeah, exactly. So it's a limited a limited range view, just just like a human would be. So it doesn't it will go back, probably a little farther than a human would by itself. Turn my neck, I can get about 90 degrees. But it doesn't go much further than that. So we have we also do a full 180 either. And so it's about it's about 150 degrees. So that's pretty harsh. And, you know, basically looking over your shoulder, right, you can look back and see what's going on. Yeah. And then And then, you know, down basically similar to a human right, you can sit down touch your head. And then the other problem is, you know, looking, looking up, you've got the back of the the head,

A.N. (21:18)

right, the Yeah, because the cameras are just underneath the chin, in a sense. So it really does like to look up. Yeah, to clear the chin. Right. Yep. Yeah, exactly. So,

J.R. (21:29)

yeah. So that's it fully, fully articulating in all of the directions.

A.N. (21:32)

That's two degrees of freedom. Can it turn sideways? I think it probably can. Right?

J.R. (21:37)

Yeah. Yeah. Being able to kind of, yeah, tilt the head, like, one of our Yeah, and that's one of the one of the most favorite ones that we have, you know, especially when we're interacting with kids, or they want to play with the robot. And so the robot kinda like a dog, like, you know, caulk its head and look at you funny. So that's one of the one of the things that we love doing with people. Yeah. Like, and then, and then a couple sounds that we put into the robot. So for them, they sound in some of the videos, which is interesting. I mean, Pixar did a wonderful job. So

A.N. (22:10)

we'll leverage a lot of their things.

J.R. (22:12)

Yeah, absolutely. Yeah. But it's, it's pretty fun. One of the in its, you know, with the sounds, people ask, you know, if it can talk in, in some language, you know, English or whatever? And the answer is, well, yeah, tech, technologically, that's pretty simple to do. Yeah, but we've made a decision to not and again, the reason for that is when you don't expect a dog to talk back to you, you know, in in a language, but it will bark and you can you can understand with with nonverbal communication, what is going on. And so in the same way, as we, you know, work with different robots in urban spaces, and they would talk to you in English, it doesn't matter how nice the robot is trying to say something in English. It comes across route, you know, you don't expect a device to be telling you to get out of the way, right. It's like, well, that's, that's kind of, versus a robot that is just making sounds, you know, more RTD to ask, then, than anything else, right? So that you, you understand, like, Oh, this is asking me to move. I was actually just having lunch today. It's at our local Mexican restaurant, and we have a robot down there that's doing deliveries all the time. And the robot had had come inside. And then there was a cup in the way where it's supposed to park. And so it's looking down there looking up looking around, you know, its head like, hey, come over here and putting sounds and then just somebody that was at the restaurant looked at it was like, oh, it's trying to tell me something went over, take the cup, got it out of the way. And then the robot was able to say thank you turn around and back into its spot. So one of your robots you're saying? Yeah, yeah, one of one of the ones like this. Yeah. So So with Yeah, so it a nonverbal communication, and that human robot interaction pieces, again, a lot of stuff that we've we've studied. So it's where we put a lot of time and effort into

A.N. (24:08)

Yeah, it seems like a thing of setting expectations. Because I think if something talks, we expect that it can really talk. Right? And then it's frustrating when it immediately doesn't recognize you, like you could say any arbitrary complex thing, and it would have no idea unless there's someone behind the scenes like this.

J.R. (24:29)

Right? Yeah, like a web chat. Yeah. You know, mean, typing, and you're like, No, this is definitely this isn't a human

A.N. (24:35)

talking. Yeah, the robot would be similar to that. Otherwise, yeah, I've seen that quite a bit from like different human robot interaction research where they want to set the expectation low. So that they can not over promise.

J.R. (24:52)

Right. And obviously, you know, other companies have done a good job of overcoming that barrier by having an actual human you know, via like a computer screen on the other side. And then you're talking face to face with a human right? The robot just happens to be, you know, the carrier. It's the media of the iPad, for sure. Yeah, exactly.

A.N. (25:10)

So do you. So how much of the robots control? Is a human in the loop? Like tele operating the robot? And how much is autonomous? At this point?

J.R. (25:26)

Yeah, at this point, the, the autonomous piece is almost strictly for navigation. So now that part's out of the way, one of the things obviously, the the software guys, they're really excited about the HDRI components automatically know facial recognition, somebody happy, sad, how do I some emotion or act? Yeah, yeah, putting them putting emotion. But you know, while it's on the job, if you will, you know, especially with disparity vision, and making sure that, you know, collision detection and avoidance and things like that is all working, it's going down a path to making sure it's not going to run into something, but at the same time, it comes into an obstacle, you know, it hits an obstacle, not hits an obstacle, it sees an obstacle stops, and then can look around. And it's kind of like, you know, looks like it's looking at flowers or trees are cars that are passing by. So that, so that is more natural than, you know, a robot that just drives you know, gets to a point and then just kind of sits there for Yeah, so. So all of the the human interaction pieces have been with tele operating, you know, having somebody on the other end, being able to emote, and it's been really fun to, again, experiment in a pretty tight loop. So you can see what do people react to? And what do they want? Like, what do they want to see from this robot on a emotional or interaction basis? And then once they have that, you know, now that we've had that we have some really good data now, just programming right? This program, like, okay, when this than this, yeah. And then the robot can know what to do in certain situations.

A.N. (26:59)

So when so the robot will autonomously go from point A to point B. It does it uses GPS or how to do that.

J.R. (27:10)

Yeah, GPS and, and Dead Reckoning and dead reckoning. So, right? Yeah. So within the tracks itself, it, it knows where it is less slippage, like you'd say, you know, better better than

A.N. (27:22)

does it use. Is it doing? It's not generating? I think, in us talking a little bit before the actual recording started. It doesn't do mapping, right, you have to go map an area before the robot goes into that area. Is it correct?

J.R. (27:42)

Yeah, for the autonomous piece. Yeah. So when we go into a new area, yeah, we go into a new area, we have, you know, an operator that's doing that piece, you know, going through finding out where the the obstacles are stuff like that. And so here's the road crossings. Here's the dangerous battle. Yeah, don't go here. You know, kind of like kinda like the Waze app, you know. So once we have all that, all that set, then we can say, Yeah, this is a this is a free and clear sidewalk, you can go, you know, full speed. This is a danger zone slowdown. This is a road crossing. We don't, we aren't, you know, there's a lot more programming, we aren't letting the robot cross streets fine. So yeah,

A.N. (28:18)

that seems that seems terrifying. For your robot. Yeah. Yes. Yeah. So what is it when someone is going to create one of these maps for the robot? What does it involve? Do they just like a handheld stereo camera to do it like? Or do you just bring a lidar through it? And it just maps everything? Or what like, how do you how do you create this initial map?

J.R. (28:42)

Question, we actually, we actually just do it with a robot to drive the robot relocating. Mm hmm. Which does

A.N. (28:48)

generate a big 3d mouse that way, and then you annotate it.

J.R. (28:54)

Know where it's a little more complex than that. I won't, I won't try to butcher it. Because I'm not one of the engineers on the project. But yeah, but being able to go through. And then with that, being able to process all the data that we need, so we can see, here's, here's what's going on. And it also does a lot of good stuff. When we're doing bigger deployments. One of the things that people really want from our customer base is Dax is really cool. It's a really cool thing, right? It just happens to be cute, and it delivers food. So you know, what it does is kind of an accessory. So when we go to a new area, it brings a lot of goodwill to the, you know, to the people that are getting in as well. You know, we're driving a robot on every path that we can think of, you know, for delivery area, and people are stopping us all the time asking questions, this robots like oh, we're partnering with this person and and so and then also gives us some good information as well for the other connectivity, you know, LTE and stuff like that. So when operators have to jump in, we got to make sure that we have what carrier do we need to use for the different agencies? What bands right Yeah, cuz everywhere. Everyone's A bit different. And so obviously, what will work in here in Oregon is not necessarily the best carrier down in Coronado, California, which is may not be the best carrier in Atlanta, Georgia.

A.N. (30:11)

Yeah. Gotcha. Okay, yeah. So you have a map of the environment that you get from driving the robot through it. And then you have the robot navigate autonomously, from point A to point B using kind of like, somewhat labeled map in some way where someone says, this is a crosswalk, this is a sidewalk, these kinds of things, then you get to your destination, and the robot is taken over by someone who's controlling it with a teleoperation setup to actually do the delivery to it. Correct?

J.R. (30:49)

Correct. Yeah. So that at this point, knowing that, you know, that's the interaction. It's not just the delivery service, right. It's, it's the experience of having a robot that is being nice to you. Yeah, so that's exactly that's where a human takes over and is able to be interactive. Yeah.

A.N. (31:05)

And so for that part, what does it look like, for from the humans perspective? Are they sitting on some sort of, like, video call kind of thing that sees what the robot sees? And they have a bunch of buttons in front of them? Where they can say nod head like this, make this noise? Like this kind of thing for picking the behaviors? Or how does it work?

J.R. (31:25)

Yeah, a little more simple than that. A couple joysticks, you know, that have, you know, a lot of lot of pre programs into the joysticks, you know, kind of your over your overview map, you can kind of see where the robots were at, you know, you can monitor several robots at a time. You know, making sure they're staying on path. And then if one needs help, you know, the operator is able to go look into it and see what they need to do navigate around or end of destination, right. Being able to, to contact the person. Let them know that they're their personal copy that's funny for the delivery, like Uber Eats or DoorDash or something. Yeah, when you have to call them and be like, where

A.N. (32:02)

are you at? Yeah. Yep. That's funny. So, okay. Let's see. So then, so they're using a simple interface to control the robot while they're doing that interaction. You were saying that eventually, maybe you could automate this kind of thing? Or you or at least part of the interaction now that you're generating data? How and then do you have the customer like rate how enjoyable it was? Or different metrics of it. So you can see how the interaction was and what parts of it they liked and disliked and whatever?

J.R. (32:40)

Yeah, good question. So we, so there's not like a Dax app? For example, people aren't ordering food through

A.N. (32:49)

us to get not Yeah, actually.

J.R. (32:51)

Yeah. Well, it's, it's, it's an interesting paradigm. We work with the restaurant and so we'd rather do integration with people so that it's their brand. We just happen to be the Korea robot that's delivered. Yep. Yeah, kinda like, you know, I lived in Seattle and Boeing, obviously the a great company. But, you know, a Boeing airplane could have Alaska Airlines on the side, or it could have delta or United or whomever, right. So. So when I go to get on my plane, I'm not saying hey, I'm on a but am I on a Boeing? I'm no, I'm, I'm with delta this time, or Alaska or whomever? Right? So yeah, it's kind of the same, you know, kind of feel. But, but with that, yeah, so especially here in our local town, it's a small town. And so being able to, you know, we've got their phone number, and then we can ask, you know, direct conversations, and see what people like and don't like we we did a test at a retirement community as well. I mean, the assumptions that we had were kids, we know kids like robots, which I think back to when I was a little kid, and I've got a I've got a eight year old and a two year old, and to them growing up. Oh, yeah. Robots around the street. Like it's just kind of a normal thing, right? Yeah. But so I'd imagine kids probably are good with it. And they do they love Dax, people, our age are, are a lot more amenable to Dax as well, because it's cool technology. So we didn't know how it was going to be received with people in their 90s and, shockingly, Dax were all for overseas. Yeah. Oh, yeah. Again, because it's not it's not a machine to them. It's it's an entity. Right? So when you watch Star Wars, for example, are 2d to into machine. He is but you don't perceive him? Yeah, you see, you perceive him as an as an entity, you know, somebody that you are having conversation with, not just, you know, directing to do a thing so so in the same way when people see Dax, you know, and it's funny, I do the same thing. I know what's behind you know, I know you know, all the inner workings of the robot, but I'll see him on the street while I'm, you know, going to get a taco. And I'm like, Hey, Dad, I'll say hi. And it's so funny. And obviously I know you know, it, doesn't it You know, I'm saying but it's, it's definitely an experience until somebody experiences or sees the difference between, you know, just you know, a cart that can deliver something to you versus having an interaction with a beam. That is a machine. It's starkly different.

A.N. (35:17)

Now. So you guys are deploying in your, in your town. So you're in Portland, right? I haven't been to No,

J.R. (35:24)

not poor. Oh, where are you? Yeah, not Portland. We're. So Oregon State University Corvallis, which is about an hour and a half south. And we actually are in a small suburb of that in a little town called fellowmen. But yeah, so not Portland. Not Not a huge Metro.

A.N. (35:41)

But yeah, small town. How many people to the town? Oh, about 5000. Wow. So

J.R. (35:46)

small. It's crazy. Small town. Yeah. Small town. I moved down here from Seattle. It's a big change. It was it was good. Yeah. Yeah. I grew up in it. I grew up in Montana personally. So it wasn't changed to go back. Yeah. Back to a small town. Right. Yeah, we've got robots now that are in Coronado, doing food delivery. Down there. That's been fun. We got some robots that are shipping out to Georgia. Cool. We next week Expo Hell yeah. So we're getting packaged up and shipped to Georgia. And then different deployments. I think Long Beach. We've got some robots in Long Beach right now. Yeah. Deliver stuff. And then. Yeah, and more people are contacting us all the time. Oh, yeah. Yeah,

A.N. (36:26)

how many? How many robots are deployed? Right now? Would you estimate?

J.R. (36:32)

What really, I could probably do the math in my head. It's not a lot. So we've got the deployment in Coronado, Long Beach,

A.N. (36:40)

we definitely use it on the order. Georgia, or is it on the order of like,

J.R. (36:44)

yeah, that's about it's about, we've got, you know, got more robots in that that are coming off the production line, but we've got actually in the field working. We've got about

A.N. (36:54)

about 10 Yeah, 910. Gotcha. They're doing stuff. Okay. Interesting. How large are you guys as a company? Like how many people are involved?

J.R. (37:06)

Yeah, so we've got there's a 18 of us, Hey, we've got the manufacturing. So we got the manufacturing side. That is, you know, getting robots, you know, off the line, then obviously, the operation side and servicing, you know, like, getting data on that, you know, how long do you have to run a robot before the tread needs to get replaced? Yeah. How many miles on their tires, you

A.N. (37:27)

know, imagine where you were out fairly fast, too, which is interesting,

J.R. (37:31)

surprisingly, not really, actually have lasted a long time, we finally we finally got one that we're like, Hey, this looks like it's certain to how many were out. We should swap this out. I don't actually this actually just happened. I think it was either earlier this week or late last week. And so I asked her Operations Manager, like figure out how many hours we put on this. Yeah, we can put that in our service manual. But yeah. And then obviously, they the engineering team, you know, mechanical and electrical and software and firmware, and yeah, so Yeah.

A.N. (38:03)

Gotcha. Cool. And how are you guys funded? For this kind of thing? Have you been accepting investment? Or like, what kind of where are you at in terms of funding?

J.R. (38:15)

It's good question. So um, so in 2002, brothers, I'll tell you the story of the founder. So two brothers, they started an internet an ISP. dead broke. I mean, they, they were volunteer firefighters lived at the fire station and paid themselves $50 A week, a month kind of thing, you know, crazy. barely getting by got it got it up and running. And then in 2011, or so kind of limping along there. Since they were firefighters. One of the brothers had this idea because they went to a fire call, they got to the fire station. And the fire trucks were coming back from the call, by the time they got there. He's like, this is stupid. Like, we have the iPhone. I'm gonna build, I'm gonna build an app. And so he built this app called active 911. And last I heard it's about 40% of firefighters in North America use this app called active. Wow. So it's done very, very well. And by his that Joseph and Joseph goal was always to have this r&d lab that I talked about. So in 2015, they started Novo dynamics. And then Dax was kind of the first big project that they said, Okay, five years from now, I think robots doing delivery is going to be a thing. So let's start working towards that end, and more importantly, working towards the future that people want to see, you know, when, you know, again, when people are thinking the future of electric cars, electric cars have been around, you know, since the 80s. Yeah, you know, long people have tried to do electric cars. But it wasn't until Tesla before now, it's something that people want, right? People are like, Oh, that's the vision of the future that we want to see not these weird box things that are glorified golf carts. Yeah, right. So in the same way, you know, with the robot is solving that, you know, what kind of thing do we want in pedestrian spaces? That isn't going to be creepy, that people are going to want to attack themselves to they can do helpful things for people. So that was that was the idea. So the r&d portion is all been self funded, you know, because these other two companies that have done well, it's been self funded up until about a month ago. And so about a month ago, I was actually, you know, January when we said, we need to, we need to commercialize this thing, what do we need to do we need to get investment. Because, you know, building you know, 20 robots is one thing, building a fleet of them. Hundreds is a totally different ballgame. So. So it's actually just about a month ago that we opened up a crowd, crowdfunding campaign through start engine, which is the intermediary, so so it's kicked off. And this first time I've, you know, I've been involved in lots of companies started a couple myself, but I've never gone after investors. So it's a whole new experience for me. You mean crowdfunding specifically? Or? Or even venture capital? You know, filing with the SEC? Yeah. Never. I've always been bootstrapped. Yep. Right. We've always bootstrapped company. So this is the first time going after investors. And it's really interesting, because the the majority of people that don't know us, you know, they don't know who we are. And they don't understand that, that the two brothers like, people that know the two brothers, a lot of people that know them have invested quite a bit because they paid another venture. Yeah. Yeah. And yeah, and they, they seen the robot, obviously, our hometown, couldn't ask for a better reception, you know, from the investing standpoint, because they're like, No, this is our robot. This is our hometown, and, you know, friendly neighborhood robot that we're investing in. But people that don't know us, that have invested. You know, we're saying over $1,000, you know, $1,000 or more, yeah, I've been reaching out to those people. And predominantly, they're engineers, you know, it's people that, that see what we're doing. And they say, Oh, this is cool. Like, this is a cool version or vision of the future that you guys have. And we want to be a part of that. So yeah, so that's where learning, you know, how do we how do we get that message out to more people? So not just engineers, but everyone that's it seems to be? It seems to be interesting that that's the community that is really liking what we're doing.

A.N. (42:06)

So yeah, well, they're interested. I don't know, as an engineer, I'm interested in robots. And I imagine it's true for a lot of other engineers.

J.R. (42:14)

It's true anyway. And it's not just you know, robotics into you know, every discipline you can imagine civil engineers are like, wow, this is. So it's kind of cool.

A.N. (42:22)

So, how was the crowdfunding? What are your thoughts on crowdfunding after? So is it still alive? Or yeah, it's still alive. So at the time of this interview, it is how long will it stay live until? Because it may be I said, it'll be like three weeks before this is published. So

J.R. (42:38)

yeah, so the the, the raise is going through either you hit the the first goal that, you know, we can hit with what we filed with the SEC. So we're looking for a million dollars by the end of the year is kind of the the goal and the target that we have, we're about a third of the way there, you know, after been going for about a month, so it will either will either hit that that mark, and the fund will shut off, or I believe it goes until I think it's a six month campaign. So it would go until until April of next year.

A.N. (43:11)

Oh, hell yeah. Okay, good. So this episode will air while it is still going if it goes to the six month time. Awesome. Yeah. What have you thought of crowdfunding? It's an interesting model. Does each person get some equity? Or they're just supporting you? Or they like get a tax robot? Or how does it work?

J.R. (43:31)

Yeah, so there, yeah, there are different ways of crowdfunding? Yes. You know, we're a nonprofit or something, it's not equity, what we're doing is actually is actually equity. So you're actually getting stuck in the company, which is, is kind of cool. So I've heard horror stories of that,

A.N. (43:47)

where you have funding crowdfunding, where you give them equity, because like, you'll have to go and get permission from everyone for all the decisions and eventually, just like veure bureaucracy weighs down the company no longer gets anywhere. That I guess you have clear about their terms kind of thing, if you guess.

J.R. (44:08)

Yeah, and everybody so yeah, start engine is the platform that we're using. And part of it is that yes, you do get common, you know, common stock, you know, which is you have shares in the company. But you also are essentially signing over your vote to the CEO, which is me. So, so when anybody says yes, you know, I'm going to put $1,000 into Dax and get, you know, however many shares they're also saying that that, okay, I know that you run minority shareholder, yep. Right. So I am your representative. So I am the vote for the shareholder. So obviously, if people want to talk to me, obviously and let their voice be heard, you know, kind of like a representative government, you know, cool and Okay, I like that a good way better talk to me. Yeah. And then same thing where it's exactly cuz otherwise, you know, your shareholders meetings. Everybody's there by proxy, and I'm the proxy and I'm the boat so, so it doesn't really bog us down administrate If that makes sense, yes, the cap table is, is big. And usually people that don't understand technology very well and the liberalization,

A.N. (45:09)

which is where the monopolization goes, right? Like or who owns what? That's typically correct,

J.R. (45:13)

right, right. Yeah. Who owns what? Yeah. So if you buy, you know, 50 shares in index bought, your name goes on the list of like, okay, you are an investor. So Oh, yeah.

A.N. (45:24)

Okay. That's crazy. Yeah, that's, I will at least we have like, you can automate everything. So you don't really have to do too much. Probably.

J.R. (45:32)

No, that's, that's all gonna say, yeah, there's not and and people that don't know technology very well, they have in their mind that you're looking, you know, that many stamps, you know, envelopes, as you're sending out like, requests, you know, to shareholders and things like that, like, No, you put it in a CSV file, and boom, everybody gets a mailer.

A.N. (45:50)

So how, how much? How much? So you said, your goal is a million, you're a third of the way there, it's been a month, you have five more months or so? How many, so in that third of a million dollars, which you've collected? How many people are involved in that?

J.R. (46:10)

A little A little over 200. So,

A.N. (46:14)

okay, so it's mostly like at least investing, at least investing like 1500 per person, on average kind of thing.

J.R. (46:21)

That's, that's the average. And we had to, we had to have a minimum that was above 100. And so what we wanted to do, and back back to the large cap table, right, the number of people that would be on this, this list of investors, people were saying, you know, don't make it too big, but at the same time, we have a lot of fans. We call it the DAX fam. You know, and we wanted, we wanted people to be a part. And so we said, Okay, we need to have a minimum, that minimum entries 170 bucks, so that people can be, you know, an owner of DAX but why

A.N. (46:54)

is that arbitrary?

J.R. (46:55)

Weaver?

A.N. (46:56)

Why $170?

J.R. (46:58)

Yeah, so we I like round numbers personally. And so our, it's 169 50. And so our with our current valuation and what the stock price is, and ends up being 50 shares, we have to we had to have a number over $100. Okay, and so we said, Okay, well, then 50 shares, so 50 shares is 169 5339 shares. Oh, I see. Okay, so that's how it works out simple. Yeah. So I just made it that was it was arbitrary, but it was a, okay, I have 50 shares of dexpot. Right, that became kind of the lowest investment that you could do, and then up to to answer your question. So up to $50,000 is, is, you know, 50,000. and above, if somebody wants to invest at that level, we said, yeah, we'll fly back to your location, and we'll have them deliver something to you. Cool. So we have nobody's, nobody's taking us up on the 50,000 yet, but we've had had a few investments that invested down, you know, several $1,000. And the average is about 1500. But I think there's a lot of investors, the majority of the investors, if you will, are about $1,000, guys, so it's been a good reception. And because it's I don't know if I know this correctly, but this is my impression, and you can correct me, because it's relatively small amounts of money that people are investing. They don't have to meet the bar, the people investing don't have to meet the bar for

A.N. (48:15)

being like an angel investor or something, which I think is like a million in assets or assets or something, and over some amount per year. So anyone can invest in this kind of thing.

J.R. (48:27)

Yep, exactly. And, and that was a law. And I might get it wrong. But it was, you know, fairly recently, I think, when 10 years or maybe 15 years when the law came out. Before that. You're right. Everything was an accredited investor, which, like you said, massive amounts of either high income, high disposable income or huge net worth that is in your house. Yeah. And so it's very, very rare air to be an angel. Yeah, exactly. And so, so with that, they said, okay, there are some rules around it. So for example, if you go to start engine, and you say I want to be an investor, they will ask you some questions about your income and things like to confirm because there are limits. Okay. Yeah. And to make sure that, that if you don't have a lot of disposable income, you're not doing it going to hurt yourself. Right. That's good. Exactly. So yeah, and there are certain limits and things again, depending on income and assets and things like that, so it doesn't hurt anybody, but at the same time people can invest in in cool startups like this totally. And it's kind of

A.N. (49:25)

early to, yeah, yeah,

J.R. (49:27)

absolutely. So they can say oh, you know, people a lot of people like I don't want to put my people that are okay putting money into crypto are usually okay putting money into, into into startups, right? They you know, it's not people they just want bonds, right? Yeah, I don't want bonds and put it in that row massively and I want to be something cool. So exactly. I get it more risk. Yeah. Yeah. Yeah. So and again, it's not you know, for a lot of people you know, $1,000 is not like a it's Yeah, doesn't exactly doesn't break the bank but if it if it goes big like that, pool. And again, it's again more a part of a, you know, I've met people that invested in Microsoft when it was nothing, you know, kind of thing. And it's like, yeah, wow, you're probably enjoying that.

A.N. (50:10)

So, so one thing that I would be curious about if I was to invest in this is about future rounds of funding, because it might be possible that they get diluted through additional funding. How does that look for like, yeah, just talk about that, because I don't know too much about it.

J.R. (50:29)

Absolutely. Yeah. I didn't either, you know, till a few months ago, when I really had to learn a lot. So the way that it works is that the goal when you go to each investment round is to the value of the company wants to go up, right? That's, that's the goal. So if the value of the company goes up, so does the amount of dollars you know, easy math, if you invested $5, and then the company, the value of the company doubles, then your $5 turns into $10. Right? Just, you know, hypothetically, yeah. Okay, value goes up. But then right, then you say, Okay, we're gonna go after other investors. So we're gonna add more shares to this. Yeah, well, exactly. It's just gonna dilute your shares, right. So your your share was $5. Now, it's $10. Because we doubled it, but now we've put more shares in the mix. So now my shares are worth eight. But it allows it allows more fuel for more growth, if that makes sense. Yeah. And the hope is that it grows again. So then that eight turns into 16. And the company doubled again, right? So that's, that's essentially how that works. And that's a very simple, very, very soon. Yeah, but

A.N. (51:34)

the dynamic, is there. Yeah. And you just hope that it's not diluted to such an extent that you make quite a bit at the end. If a company does have some sort of liquidation event?

J.R. (51:46)

Yep. Yep. And that's the goal. And I've, you know, talking to venture capitalists and angels, and just kind of getting a sense for what what are you looking for, you know, what is what does an angel investor or a venture capitalist looking for?

A.N. (51:58)

What do they think?

J.R. (51:59)

Most? Yeah. Most of

A.N. (52:03)

them

J.R. (52:05)

have no experience with at least the ones that I know. I've talked to most of them have no experience with it, because they haven't had to write they they're that rare air of angel investors accredited investors. So are they don't need to have an investment firm or something. Exactly. This kind of Yep. throwing around the money. A lot of people. Yeah. And since again, because it's a fairly new concept in the scope of, you know, the history of the world, it's a fairly new thing that you can get into these crowdfunding rounds. They don't have a whole lot of experience with it, but a lot of them, it seems that the tides are turning a little bit. I'll give an example. So there's a, an angel group are called the Oregon rain. And it's more for small, small, small startups. You know, like, you know, I need $30,000 to get this, this idea off the ground. Yeah. And so because they're angels, and trying to figure out how they're going to do it, they actually partnered with another intermediary, I can't remember, but basically who it was, but essentially, this other intermediary, like start engine was going to give them a white label version of what does it mean, programmable? That version? I don't know. Yeah. So yeah, so a white label version would be. So for example, if I have a software, let's call it Microsoft Office or Microsoft Word, right? So a total a white liberal version amendment.

A.N. (53:22)

Yeah.

J.R. (53:24)

Okay, so I can't just I can't just say, hey, it's Microsoft Word. Yeah. But if I had a white label version, if you came to me and said, I'm going to give you you know, $300 billion. So I have a white labeled version. It's not Microsoft Word anymore. It's now dexpot word, right? And so I the label, or you buy, essentially, okay, you buy it, you buy the rights of the software, but they put your name on it. And so that's so a white middle version of startengine would be, you know, Oregon rain.com/investors. But the actual software was not written by Oh, yes. Investors, okay, by another company. Yeah, that makes sense. So, so, so I, it looks like it's becoming more. There's more awareness around crowd fund so that in these angel groups seems like,

A.N. (54:08)

okay, that's what you were. So I was a bit lost for a sec. You were saying basically, that's an example of crowdfunding going into something that is a bit larger. Yes. With investors or something?

J.R. (54:20)

Right. Yeah. The angels. They are they know of it, but they haven't participated. But it seems to be just the research that I've been doing, it seems to be that they're more open and receptive to it. Interesting, more, because you have companies that have gone through crowdfunding rounds, and they've they've done they've done it successfully, meaning that they, they got the money, and then they executed on the goal correctly. They started getting revenues. And now they have revenue. So now they could go into like an angel investment group or venture capitalists and say, here's our revenues. And here's the funding that we got before. Now, would you you know, invest in this and there Now seeing because they're seeing this happening more often, they're more privy to the fact that, oh, we might want to look into investing in crowdfunding options ourselves.

A.N. (55:09)

Why would so why would a company go into crowdfunding as opposed to seeking investment from, I don't know, like a, like a pre seed or some sort of company or some sort of angel investor or something like this, because I understand that one of the benefits of having investors come together for your company is that they can help you make while they can connect you to the right people, or they can mentor you. So you avoid potholes. Why would you choose crowdfunding over a specific like one larger investor? Kind of thing?

J.R. (55:48)

Yeah. Yeah, to two main reasons. So January, then back January of this year, when we were looking at going after investment, reached out to some venture capitalists and friends of friends that we have, but no people. And essentially all of them said, Okay, well, what are your revenues? Zero. Okay, so if your revenues are zero, that's a pretty big gamble. Right? You're for them. That's the perception market event

A.N. (56:19)

or these kinds of things. So neurons,

J.R. (56:21)

okay. Yeah. Yeah. So I did at this point, they would look at that and say, Oh, it looks like it's still like a research phase. It's like, Yeah, but we want to go from a research phase into a commercialized phase. So in order to make that jump, it's going to require money. And so I was talking to one, one venture capitalist who he been on the other side of the table to write he had he had startups gone after venture capital and had some successful exits. And, and he said, he said, what you're going after right now is he called it dumb money, for that reason, right? Because when you go after venture capitalists, you're not only you're getting their Rolodex, right, the old Rolodex is where people had contacts and things like that. So you're going after their Rolodex as much as you are on the cash. And so that's kind of the next step that I see.

A.N. (57:06)

And also just advising to not just rollback, but like they can really no, because they could have could have seen people go through similar stages. And they can be like, Ah, this Yep, like, at this point, you should be working on this. And at this point, you do this, and this is how we get to the next stage, our Series A or whatever it might be.

J.R. (57:25)

Yep. So yeah. So with with the, so as we talked to different people, essentially, it was, you know, that was kind of the feedback we got from several, the, you know, get some revenue didn't come back to us kind of thing. So we looked at that and said, Okay, well, then crowdfunding seems like, a good option for a few reasons. One was because we everywhere DAX goes a celebrity. You know, everybody, kind of everybody, there's, about half of the people will kind of walk by like, oh, yeah, I see robots trying to talk to me every day, you know, like, and I might say, I'm trying to

A.N. (57:55)

go, where is this? I don't know. I feel like I should be in the area with all the robots. And you were mentioning that it's not legal for robots to drive on the streets here. But I'm in I'm supposed to be in the tech area. And I so seldom see robots anywhere doing anything. Work From Home stuff, but yes. Okay.

J.R. (58:16)

Right. And then and then the other half of people are, you know, getting out of their cart. We haven't caused any, you know, auto accidents from people like rubbernecking, but it is fun being you know, walking. It's fun walking behind dact. Yep. And seeing people on the road because people will look and even if they're having a bad day, it's funny watching the people in the car because they will, they will turn their head look at the robot. And you know, the look at DAX and then everybody smiles. Where has this look of like astonishment? Yeah, what is that? Did I just see that? But everybody smiles like, well, that is that's cool. You know, like, huh, I just saw something special today. So anywhere, the DAX goes. He's a celebrity. Like I talked about, you know, the hometown film, if we wanted to be able to say, hey, you can invest to for crowdfunding. So yeah, like, give you that opportunity. The other piece of crowdfunding is it's a lot of, it's a lot of marketing. Right? Film is, is a, it's a great little small town. But it's a small town. Down here for me. Yeah, versus Seattle. And you got 4 million, you know, not as big as you know, LA where you're coming from but, but need more eyeballs on DAX, right, more the more eyeballs, the C DAX, the more people that have invested interest literally, right? If somebody is in Texas, and they put $1,000 into DAX, they're probably also more privy to be like, hey, I want DAX down here, too, you know, or what connections do you have any larger sense? That makes sense? So those were the two big reasons that we said yeah, let's let's go after crowdfunding and make it work.

A.N. (59:39)

So what do you think? So I coming from, so I was in grad school doing human robot interaction for a little bit. Oh, yeah. And one of the things that was very concerning for the community was kind of the this novelty effect. So like someone can be, like really excited about the robot, and they'll like, do all the things the robot wants and behave really well around it. But then over time, so as weeks go by, and months go by any, like, maybe the robot becomes less influential on those people, because they now are like, I know what it does. I'm kind of sick of it. This kind of thing. Yeah. Any any thoughts on this? Because the robot is currently a celebrity. But I wonder, could it be a novelty effect? Or is it? How do you think of these things?

J.R. (1:00:35)

Yeah. So if if all the DAX did was be novel? I would agree with that. Right. So differing things and so right. So do you have to have a useful function as well? Right. So if he's, if it is, same with you know, like, I would say that, that if it's doing useful things for people, whether it's, you know, delivering you food or, you know, patrolling grounds, like a security robot or something like that, right, if it's doing something useful, it's not only novel and cool. You wouldn't get sick, because it's actually helping. It's justifying it feels way. Yeah, exactly. Especially in the delivery space, because delivery. You know, if you look back to the 1970s, when they started tracking it, called the labor force participation rate. And so the US government tracking, it was like, US government. Yeah. Department of Labor. Okay. Yeah. The Department of Labor is looking at how many people are working compared to the population, and that that number has been in steady decline since the 70s. COVID exasperated that. Yeah. But you know, where was less people working in at the same time? If you look at just food delivery service, in general, that's been on an increased, you know, exponential growth path for the last several years. COVID exasperated that as well. So the restaurants and the people that we talked to in the delivery space, it's almost an expectation. Now, you know, where, what if you own a restaurant delivery, you really want a restaurant? Yeah, exactly, exactly. You might not, but Uber Eats or GrubHub or DoorDash, or somebody is going to deliver this food to me, right? It's kind of the expectation. Yep. And so that's, maybe I don't want to say an unrealistic expectation. But when you don't have enough people to fill those roles, it's just not going to happen. So for example, in Coronado, one of our guys walked into one of the restaurants in the back and they said, do you see all those orders, and there's like 20 delivery orders sitting there, like they, they've been sitting for hours, and there's nobody going to come pick them up, like very doubtful that somebody is actually going to get on UberEATS or GrubHub, or wherever, and go and pick those orders up and take them to the people that want them. You know, last delivery experience I had, when I was on a trip to Montana, is we had delivery set for 1130. For lunch for there's only five of us in the room. So okay, lunch is gonna be here at 1130. We we ordered it ahead of time, make sure we got into delivery schedule, and it didn't show up till 2pm. And that's just because of you know, there's not enough people to fill those types of roles. Is that

A.N. (1:03:04)

a thing that's more particular to small towns, do you think like basically, is the whole like, say, middle of America that would have smaller towns? or basically any, because I'm thinking of like, I'm in San Francisco now. And generally, it seems like the food delivery services are quite quick. But maybe this is so it hasn't been in my experience, as you've described, where there's like, 20 orders for this kind of thing. Is this particular 20 locations, I guess, or any types of sounds or types of towns?

J.R. (1:03:42)

Yeah, that's a good question. So Coronado, California isn't isn't exactly like, you know. Yeah, yeah. It's right outside San Diego. But it's right outside San Diego, right, you got to go over the bridge. So it's also very wealthy zip code. And so if you are living in Coronado, you're probably not the person doing the live delivery. So people have to come in, and then do the delivery within that, that area. So, so that's kind of unique thing. The other thing is that a lot of people, you know, door Dashers and people that do GrubHub and UberEATS, and stuff like that. A lot of them don't want to do the short, the short stuff. Yes. Similar to like, your Uber driver wants to drive you 30 miles away to an airport, they don't want to drive you three blocks, right? Because they set

A.N. (1:04:27)

up for but in general, some of them are trying to generate trips and these kinds of things.

J.R. (1:04:32)

Right, yeah. But in general, you know, you're you're making more, if you will, you know, the longer you have. Yeah, you're doing the long trips. So so same way you know, for robot you know, this you know, the DAX is not DAX is not optimized to go 30 miles you know, he's he's got a pretty good like, one, one and a half mile range. And those are that's a good niche to fill.

A.N. (1:04:56)

Mm hmm. Okay, could you Imagine pairing with vehicles and things where this are like kind of like the package delivery idea where it'd be an autonomous car that drives the robot and the robot does the last little bit of the delivery. Could you imagine that being used in that kind of thing?

J.R. (1:05:16)

I could. It would be it would definitely be interesting. Obviously, there's a lot of people that are working towards those types of solutions. Yeah.

A.N. (1:05:25)

Even in the Oregon area to thinking of agility robotics, I believe. Yeah, they're trying to do the legendary

J.R. (1:05:31)

robotics thing. They're just up the road from us. So Oh, cool. Yeah. Yeah. Hello,

A.N. (1:05:37)

litres of package delivery, or whatever? Is that? Yeah.

J.R. (1:05:41)

Yeah. And it's a very interesting it, you know, that, you know, last mile delivery is always the most expensive, you know, you look at logistics in general. That's, that's the hard stuff to figure out. Yeah. So very potentially, there's a lot of, obviously, a lot more that has to go into the logistics of that loading and unloading a robot moving parts.

A.N. (1:06:01)

You need, like a little lowering platform or something to get the robot

J.R. (1:06:05)

down there. Yeah. And we've got a band that we, you know, haul robots in. And we have that, but it from from a large scale, you know, like if you're a big Logistics Center, like Amazon, or UPS or something like that, right. It's a lot of moving parts there. And so sure, eventually, I could see something like that, like an Amazon locker, you know, that that drives to a town, but then every locker opens up and becomes a little robot and drives to your doorstep. You know, it'd be super cool. Yeah. You know, there's, there's probably something there. But there's a lot of not just not just, you know, technological hurdles to overcome, but a lot of like legislative hurdles to overcome as well.

A.N. (1:06:43)

Which I do want to talk about for sure. Yeah, legislative things are very interesting. But just before we do talk about that, when so you're not generating revenue yet? I believe, right.

J.R. (1:06:57)

We are we are now you're just doing the stages. Yep. Gotcha.

A.N. (1:07:01)

I'm just wondering, What will your business model be? Or what is it now? Yeah, for generating revenue? And maybe you'll flip it around a bit. But

J.R. (1:07:12)

yeah, um, so the, the thing that we have right now is kind of twofold. One is that we realize kind of a back to the DAX as a celebrity thing. We can talk about it now, there's been a few events that the DAX has been in. One was for a TV show called the coma FD. So there was a bunch of DAX robots that they turned into firefighters and and get a bunch of promotion for a season or get one with or under. Yeah, exactly. Right. He was an avid he's like, my two worlds collided, this is amazing. And then, and then one was for a Jack, Jack Daniels. It was some some beverage company. And we did a delivery promotional event at a golf course in Napa Valley. Things like so. Yeah. So different marketing agencies are reaching out to us because it's a unique, it's unique enough, right. It's novelty enough at this point, that it's a great marketing engine for them. So so being able to do appearances and stuff has been fun. And obviously, it gets our brand out there as well. So that's fun. That's that's been generating some revenue. The the flash entertainment robot, yeah, right, exactly. But those are, those are gig events, if you will, right. It's not recurring revenue. So but the the actual business model is revolving around that you have different deployments in different states that are, you know, renting robots for, you know, 1000s of dollars a month for robot, and they're gonna do deliveries a service for you, which also keeps it so that we can take care of service and those types of things. And then those that's the monthly recurring model that we want to spin up. So Gotcha.

A.N. (1:08:47)

So it's been where they buy the hardware, or where they rent the hardware. They

J.R. (1:08:53)

rent the hardware, correct? Yeah, yeah. They're renting the service of delivery. And we happen instead of us being humans there. We have robots that do it for

A.N. (1:09:01)

us. Gotcha. So I just recently, at the time of recording this interviewed cobalt robotics, and they're doing something very similar for security guard robots, where they're having the robots basically be the legs and then they have someone controlling it at a very high level. And so to me, this is kind of similar to what you guys are doing, where it's you will have the robot be autonomous on the navigation and then you'll have a person control the robot for the interaction for the delivery. Is that Is that what you imagine? Yeah,

J.R. (1:09:38)

that's yeah, I imagine that that more and more pieces of that puzzle will get solved, you know, autonomously. Yeah, again, like the facial recognition pieces. Are you happy sad how do we make your day if you're a kid, what do we do you know, you know, those those types of things. But then say what the delivery piece you know, being able to get to an end destination, and then somebody comes out you realize that There's a person now in front of you, you can interact with that person. And then open your drawer if you know, you can see the code is correct, you know, those those types of things can all be young automatable. And that's the next pieces.

A.N. (1:10:12)

Gotcha. So the goal is full autonomy as much as possible. I'm just wondering, yeah, if you would have a, like, I don't know. So in cobalt, they have, like, half of the people there are just watching over the security guards. They're the people that take over autonomy. And I would wonder if at least before all this stuff is automated, would you have some sort of similar model where it's, you have a bunch of people working to do the autonomy, or do to supplement the autonomy. So when it can get there?

J.R. (1:10:45)

Yeah. In you're exactly right. It's more of the and that's what we tell people, it's semi autonomy, because when people think autonomy, they think, full on this robot is our duty to and it does exactly what I want, whenever I want. But no, the semi autonomy piece is being able to increase the number of you know, the ratio of operators to actual robots that they oversee. Right. And that's the goal is to have greater and greater autonomy, which requires fewer and fewer operators, that aren't necessarily interacting with every single person that they see. But making sure that robots are, are generally staying in line with what they need to be doing. Yeah, you know, or if they see something that's completely, you know, a tree falls in the middle of their path. And now they have no way around, because we're not letting it go outside the bounds of you know, it's on a sidewalk, it's like, we can't go into the road. You know, so not very nice to take over and make a human decision of okay, we're gonna have to find another path, right, turn the robot around, but that kind of thing. So. So there will always be an overseer, you know, somebody that's commanding, you know, a crew of 20 or so robots. So

A.N. (1:11:48)

cool. And how did you decide to do the subscription model as opposed to like, they buy the robot, and they're on their own, or they buy the robot and you like, manage it somehow? Or they buy it? And then they are a subscription thing? Or like all these permutations? How did you pick the one?

J.R. (1:12:09)

So the current one, investors really like MRR monthly recurring revenue, they like seeing that this is going to come in whether you, you know, yeah, essentially all the time. Right, versus fairly predictable. Exactly. And that that really rings true for a lot of investors. So that's one. The second is, is that when people have asked us to buy the robot outright, we've kind of played with the idea. But at the same time, we said, Okay, well, we're gonna have to come up with some sort of service contract on top of this. Because, you know, you could take a, it's not ubiquitous enough, where you could take a laptop to a computer repair shop, and they could probably figure out what to do or diagnose it. You take a robot that nobody's seen before, to like, you know, yeah, a computer repair shop, they're not gonna know what to do. And so at the same time, we said, it wouldn't, it doesn't make a lot of sense, because we did come up with a model and said, Yeah, well, you can buy the robot for, I don't remember what the number was, like 30 or $40,000. And then it's still going to be a, you know, a fiver $1,000 monthly service contract to have that robot. So we said, you know, if we're, we have to go that model anyway. And explaining that to people, they understand that they're like, oh, yeah, because if a track falls off, or you know, a tread wears out, or whatever, you know, yeah, you need to get to the restaurant to repair that. Exactly. Yeah. And so, so having a service model made a lot more sense. Yeah. To, to us, and to people that have been talking to so

A.N. (1:13:30)

yeah. Let's see, I have a few questions that I want to ask all at the same time. But what? So you're doing things where it's like, they're in Atlanta, they're in California, they're in Oregon? Does it mean you're going to have to set up technicians at all these locations so that they can do these repairs? Or how the kind of the logistics of that work?

J.R. (1:13:52)

Yeah, so it's, um, we actually what we do is we actually find partners in those areas, that we can then train on some minor stuff, you know, usually it's, you know, computer repair shops, or people that are in, you know, manufacturing, robotics type hardware, maybe not robotics itself, but you know, something that they see a circuit board and they don't freak out, you know, they it's not like taking into a bicycle repair shop. Right? Yeah. So somebody that has some experience with technology, so we partnered with them, and then then become essentially our point of contact for shipping, you know, robots in and out of a city. And then we also we keep a hot spare that's readily available inside swab. I see. Correct. So if the robot has a problem, something happens. That yeah, we can we can within the day, swap the robot out with a new robot ready to go. If it's simple enough that we can repair with remote hands. We'll have them do that on a surface contact. If it's more complex than that, you know, the one that was there's a hot spare in the crate. We saved throw this one in the crate we got another one coming your way we ship a robot and they robots basically meet, you know, somewhere in the middle of America. So

A.N. (1:14:57)

somebody then The next thing I don't know if I'm correct on this, but I imagine that food delivery is a fairly low margin industry, which means you need a lot of volume. Because and I think this and I'm not I don't have data and I'm not really aware. But I don't think that the like Uber Eats drivers are paid that well, or any of the any of the delivery services. And so I think it's more about like, How many can you get? Does this and I'm wondering basically, for how long before the robot would like pay off its existence, and start being profitable to the companies in this kind of thing? If it is a low margin thing? Because they have to make a lot of deliveries? Probably, yeah. This kind of thing? Yeah,

J.R. (1:15:51)

it really depends. So a good example of this would be one of the people that is getting the service down in the Atlanta area in Fayetteville, a company called nourish and bloom, and they're a kind of a cool new concept of a grocery store that is essentially fully autonomous with people that can help conserve eat it. Yeah, so Amazon had one that was similar in Seattle, where you walk in facial recognition, here's why. So it's

A.N. (1:16:17)

that kind of thing.

J.R. (1:16:18)

No, it's not, not quite that that degree, but similar. And so people are coming in, they can kind of self self serve, they don't need to get checked out. Right. It's, it's, it's automated, you know, okay, so they, they wanted robots as well. But if, for example, you know, the, the payload is huge in a robot. So, how do you how do you make that valuable, and it'd be similar to, you know, if you're going to shop at Walmart, for example, versus Whole Foods, you know, the amount of money that you're going to make per bag at Whole Foods is, is a much higher margin. Walmart, for example. It's a totally different market. So we're not It's not like a luxury market that we're going after. But it is more of a high end market that we're that we're

A.N. (1:17:03)

targeting, ah, so these things, evaluate your ad, like heating and cooling, and this kind of thing on your robot so that the the food arrives and is still tasty, not like, with DoorDash, if it's slightly cold, or something like this. So it's the luxury market with subsidized a bit too,

J.R. (1:17:23)

you know, if that makes sense, they're making higher margins on products that they're selling anyway, I mean, you know, Whole Foods and Walmart, they don't sell exactly the same thing. But the margins are definitely different. Right. So, so that's part of it. And then the other part is, is also the, that becomes a unique, you know, unique service that they can provide that nobody else does. So then drive more traffic from a marketing perspective. Yeah, oh, I can have a robot deliver it to me, these people can do that. And these people can't. So I'm going to shop here, even if it cost me a little bit more.

A.N. (1:17:51)

I know that it's cool for you guys. But I'm looking forward to the day that it's not uncommon for robots to deliver things. I think that would be sweet.

J.R. (1:17:59)

Yeah, me too. Actually.

A.N. (1:18:02)

That'd be nice. Okay, now, complete non sequitur. How does your robot deal with stairs?

J.R. (1:18:10)

It doesn't.

A.N. (1:18:11)

Yes, so what? So I'm imagining, like in San Francisco, it's not always the case that there are nice curbs to get around. So it may just be like a sheer drop onto the street kind of thing. Does this? I don't know. i How do you guys decide to not do stairs? With the tracks? And it? Does that affect? Where the rope? I'm sure it affects where the robot can go? And what cities you can deploy in things like this?

J.R. (1:18:49)

Yeah, the the, because there's a couple things as well. I mean, San Francisco, Seattle as well. The San Francisco especially is known for some very steep streets. Right. And so there's some areas that we were five degrees and

A.N. (1:19:03)

someplace. Yeah, exactly. Yeah. This kind of, so

J.R. (1:19:08)

we wouldn't, we wouldn't put a robot in that area just for that reason. So we had to make a couple of design decisions, right, like all engineers do. And one of those limitations is we said, Okay, where can we deploy this thing? And we said anywhere that is ADA compliant with current, you know, if

A.N. (1:19:23)

visibility American Disabilities associated to a wheelchair should be able to get

J.R. (1:19:27)

around there. Correct. So any ramp, and I don't remember what the exact slope is, for that, but we said that that that is going to be kind of the max grade that we're going to we're going to be on which is still pretty steep. I think it's 18%, if I'm not mistaken. So it's still significant. I mean, somebody is a real metric. Give me

A.N. (1:19:48)

degrees. So yeah, 100% is straight up and so 18% is 20% of 90 degrees. So

J.R. (1:19:57)

yeah, I think yeah, exactly like that. So, it's not nothing though, right? It's, it's, it's sticking really soon. So, so but you know, same thing, you know, at some point, you know, you don't want this thing flipping over on its back, you know, and, and that kind of thing. So stairs is something that we know that we'll have to deal with eventually, you know, same with elevators, especially for

A.N. (1:20:20)

delivery to Yeah, elevators, like I mean, my apartment building. There's stairs on the way in and then elevator. So it would have to meet me outside.

J.R. (1:20:32)

Correct? Yeah. And I know that, you know, a lot of UberEATS drivers and things like that can you know, come to the door and true, you know, a lot of them have to get wrong in you know, you hit a code. And usually people meet you in the lobby, something like that. So there's, it's not completely outside that the paradigm. But when we're looking at places to deploy currently, yeah, obviously, we're looking for places that are more flat, that have less stairs that are usually with up kept sidewalks. I mean, the small town that we're in is a nice little small town to live, nice place to be from. But it's a it's a great place to test because as terrible sidewalks. So we're finding out a lot of things about potholes and what we can and can't do gravel. You know, the elements getting through mud, you know, that kind of stuff, which is obviously great for a testbed, but not an ideal deployment.

A.N. (1:21:19)

Yeah, let's see. And we are starting to run out of time. So I would love to talk about legislation for the robot, because this is an interesting thing that I hadn't considered before you brought it up. So yeah, just tell me a bit about it. In January,

J.R. (1:21:35)

yep. I didn't know much about it either. Until I until we started looking at deploying into different places. And the first place that had legislation for delivery robots, was in Washington, DC, and that that came out in 2017. And had different different factors involved weight, maximum speed, when it can and can't be on the roads, you know, safety information, etc. And so different states have adapted that to very similar, you know, everything is fairly similar to that. Pennsylvania, for example, they said that, yeah, we're going to give it the same rights as a pedestrian minus a little bit, they do have to yield to pedestrians. And, and so, in California, and Oregon, interestingly enough, doesn't we don't have state laws regarding robots on sidewalks. So it's up to individual cities and municipalities to do that. San Francisco banned them in in 2017. He said, disappointment. Yeah, yeah. And, and then they made it very, they made a very small exception and a bunch of hoops that had to go through, I think Postmates is allowed to have like, two in San Francisco now. And that's it. And the reason for that is, has to do with pedestrian space, you know, people, again, the reason that dexpot is, is it's just different, you know, it's different than any other delivery devices out there. So in the legislation, they call, they call it a PDD, or a personal delivery device. So which essentially can be whatever you want it to be to put a skateboard and put a motor on it, and then say, you know, drive it remotely with a camera. That's a personal delivery device, and it can drive wherever it wants on sidewalks. And that's how this legislation refers to correct what we're referring to as a delivery robot. Okay, correct. Yep. And so the PD ds, and so a PD DS, we want it we're working very hard with we, you know, worked with the state of Maryland on their legislation working with the seat at Coronado on on rules around it. And the thing is, is that people don't have a concept of what it could be like to have DAX, if their only experience is a remote control box that is getting in their way at their feet. They don't want it there, versus DAX, which is a robot like you would see in a movie that is welcome in human space, like a dog is. And so helping, you know, change the legislation is a big deal, because there are a lot of places that it does not make sense. And it's because people don't want delivery robots on their sidewalks, you know, but when they if you just look at it from that box perspective, you know, like an engineer, like I want this box on wheels, it makes sense. It's an easy design, it's easy to do. But that is not what is being accepted. So the human acceptance piece and pedestrian space is a much bigger deal than most people realize. And that's that's the thing, the DAX is essentially solved for that, you know, being able to be welcomed on pedestrian spaces.

A.N. (1:24:33)

That's interesting. So it's, it's as if part of your role in being involved with that spot, is that you have to work with lawmakers and things to try to show them that this robot is okay in human spaces, and should be allowed and this kind

J.R. (1:24:52)

of 100% Yeah, because, because if this is the future that we want, namely want robots to be able to do more things in pedestrian spaces, we have to make sure that we're not legislating our way out of existence. And not just us. I'm talking about, you know, robotics engineers in general, if you're doing something that is in is public facing, most people think it's all magic, right? People that are engineers understand there's a lot of, you know, mechanics and software and connectivity. There's a lot behind these machines. So I guess my hope is that, you know, other people that are out there are not doing things that would then jeopardize, you know, urban robots for everybody else, similar to how drones were, you know, when drones were introduced. A lot of people did a lot of really silly things with them, because it was cool technology. And now there's so many regulations around drones. And they can be very, very useful, but they're not as useful as they could be, usually because of legislation because people did some silly things, especially in the beginning. So

A.N. (1:25:57)

okay, so how do you avoid doing silly things? For this? If you're trying to like, do something and you don't know, it'll be silly. I assume people don't go into an endeavor going, this is going to go badly. And it's going to hurt everyone, because crazy regulations will be passed for this, like, way over correct. And make sure this doesn't happen again. Oh, absolutely. So how do we avoid things?

J.R. (1:26:23)

You have to question? So like the San Francisco problem, for example, I'm sure that the engineers were like, Hey, we're gonna take this thing on the road, and everybody's going to be in love with it.

A.N. (1:26:34)

What's the problem? What do you mean, the oh, that they're,

J.R. (1:26:38)

they're illegal in San Francisco? Oh, yeah. We can't make a delivery robot to San Francisco. And whoever, you know, was a part of doing that test. The engineers, whomever, I'm sure they were thinking this is going to be amazing. And everybody's going to love it. And then people did. And so when we go to new places, for example, usually we're finding out you know, what are the rules, you have rules around this? And then giving usually the police department a heads up, you know, hey, we're going to be doing this, we're going to be here. So when they, yeah, when they get phone calls from people, hey, this robot is or there's this thing in my way, or this thing? hit my car, whatever. We haven't, hopefully not. But yeah, yeah, there are others that have. So when they get calls, they're not completely blindsided, and then they get an angry phone call from the mayor. Right? And then that's how those things happen. You know, it's a couple decisions makers, they hear from some angry people in their city that can then essentially, like, make it up make little buddy eyes. Yeah, yeah. So just more being aware and conscious of, you know, where, where are you at, you know, who's around, and making sure that that, and not necessarily calling City Hall because they won't know. But if you call usually the local police department, and just ask, Hey, you know, we're gonna be in this area with this robot. Is there any rules around that? They'll usually be able to tell you yes or no, you say, great. We just wanted to let you know we're in the area, give you the heads up. And so that, you know, if it's some robot rolling along the street, and people think Armageddon just happened, that they can realize no, no, no, this is this company, and I've got their contact phone number, you know, so.

A.N. (1:28:08)

So this is like going to, I don't know small towns and things like this and or at least individual places and talking to them and telling them. This is what we're doing talking to the local police department. How How will it be to go bigger than there? So if it's like, statewide, or federal or?

J.R. (1:28:27)

Sure, yeah, when I talk to local police department like Portland, Oregon isn't isn't a small town, right. But I'm calling you know, the Portland Police Department and talking to them, whoever you have to talk to, you know, at that, at that level, let them know what's going on. Going from there, though, most of the regulations are actually going through the Department of Transportation. So it's finding who at the department of transportation needs to be aware of this, if you're going to make a law. Again, the legislative process takes a long time, you know, for laws. So I would not recommend usually people go that route first. Especially if they're a smaller VLANs

A.N. (1:29:03)

are going to keep growing kind of unfettered, you would need these laws in place at a state and then maybe federal level, correct for this kind of thing.

J.R. (1:29:12)

Yeah. And federally, I don't know if that will ever, it might, you know, come come about some regulatory agency like the FAA, the Federal Aviation Administration for drones, but probably not, it'll probably stay at a state level and a local level. And so the more that you can build rapport there

A.N. (1:29:29)

by completing all of the local police departments, yep,

J.R. (1:29:34)

that you're safe on there, and you're going to be responsible. That's the thing that people care about more than anything, it's this thing is safe, and we're being responsible. And even the bigger piece of that, you know, obviously we don't have time talking about privacy and you know, things like that data security, this isn't big brother spying on you, you know, that kind of thing. But, but as long as people know, they know that they

A.N. (1:29:58)

may get it written in Yeah, yeah. Okay. Yeah, that's cool. Let's see. So what is the timeline for you guys like beginning to wrap up? What do you expect the next few years for you guys? Or where are you going? Yeah, yeah.

J.R. (1:30:19)

So I expect the next couple, a couple of years, you know, 2022, the goal is really to have about 100 robots out in service in different places, we want to keep it to about, you know, 10, maybe 12 geographic locations, you know, so by that, you know, Los Angeles, well, there's Long Beach, and there's different places. So, you know, if we're in Greater Los Angeles, how do we grow from there from that hub? Atlanta, Georgia, you know, how do we grow from that hub, we go to Washington, DC, how do we go from that hub. And so we want to have about 10 to 12 hubs or, you know, bigger deployments in a geographic region that we can then grow and expand from, because that will really, you know, really set the tempo for commercializing. So with the crowdfunding rounds, being able to build the robots that we need to deploy them into these areas. And then once we have that, essentially, the operational piece has been figured out, right, there's revenue, so that more it'll attract more investors, maybe higher level investors. So then we can go from that and then manufacture on a greater scale to, you know, 1000 o'clock, this kind of thing? Exactly, yeah, bring the cost down, bring the autonomy up. Because both of those things happen, right? It makes you know, better, better margins, so you can grow faster, and then be able to be more the face of robots in the same way that Tesla became kind of the face of electric cars. We think that DAX is going to be more of the face of, of urban robotics that does delivery, because delivery robots, again, isn't a unique example, or isn't a unique idea. But the way that we're going about it is that it's cool. People want that to be in their towns and on their streets. So it's going to be readily accepted in in ways that other companies probably won't be right now.

A.N. (1:31:59)

Or hopefully, so yeah. Yeah. Awesome. Okay. And then, what have been some of the, like, larger challenges for you guys so far? And what would you do differently? to kind of get even where you are still?

J.R. (1:32:14)

Yeah, um, larger challenges. You know, one of the things, you know, is, is building an operating system from the ground up because

A.N. (1:32:26)

of an operating system from the ground. Yeah.

J.R. (1:32:28)

Oh, yeah. Called DAX Oh, s that are DAX, those, say, Dax.

A.N. (1:32:33)

It's like, we're at the end of the interview. And that's a very interesting thing. Okay.

J.R. (1:32:37)

It is very interesting. The founder, great, great guy. And so his his thing is, you know, obviously, business savvy enough to know that we can't just open up access to everybody out there in open source at this point. But that is the ultimate, you know, I think that's his ultimate objective is he wants to be able to open up DAC, so S, which is a Ross like, software, you know, but it's way more than

A.N. (1:32:59)

that it does message passing and things like this store. Yeah. Yeah. And it helped and visualization perhaps.

J.R. (1:33:06)

Yeah, it does, it does a lot of pieces. And one of the pieces that we have is, as far as the firmware piece is called a nova node. So it, it basically is an extensible piece of the robot. So you can plug in another Nova node. So for example, we say, Okay, we need an iteration where DAX has an arm so we can push buttons or whatever, we plug in another Nova node. And then DAX knows, oh, now I have this arm. And now I'm able to do this, this thing. And it's, it's a higher level intelligence and has to do with a lot of the patents in the higher level architecture stuff. But Daksa was that was that's something that is absolutely fascinating to me, my background is software. So it's, it's really fun for me to be able to nerd out with the the software guys a little bit and hear the high level architecture around it. And again, that that was one of the things that the founder wants, I mean, he he, this is the vision that he has for this could be like a future thing. So we want to make a an operating system that, you know, that is able to be open source so that other people can can develop against it as well and build their own stuff similar to like how Ross is, if that makes sense. You know, people are familiar with the Ross, you know, robot operating system. But DAX DAX Oh, s would be similar in that vein.

A.N. (1:34:22)

So like, is it? Is it a real operating system? Because the robot operating system is not a real operating system? It's like a flavor of Linux or something, or what is your deck? So yeah, you'd be right about that. Yeah,

J.R. (1:34:35)

it's it's it's more of a flavor of another one. But it's, yeah. When you talk about Ross, people think oh, operating. I know. Yes. The communists know, for sure. Yeah. But so. So yeah, we'll put it that way. Yes. So yeah, on Linux, a lot of Linux based.

A.N. (1:34:54)

I feel like it's a very interesting idea. And one of the things that I see is kind of like I mean, one of the things that's really nice with Ross is the whole community and all of the things that have kind of become fairly battle hardened there. If it's if it's too big of a deviation from like Ubuntu or some of the other larger distributions, I feel like it would make it so that there is you can't reuse a lot of what exists even like, sensor packages and stuff. But you guys are clearly doing things with different sensors. So yeah, I don't know, right? Yeah. Yeah. Yeah. Is it similar enough that you can use like anything on Ubuntu? Or? Oh, okay. So it's not that large of a deviation for this kind of thing? Is it real time? So is it a real time operating system? Yes. Cool. So it's deterministic? That's awesome.

J.R. (1:35:53)

Yeah, it's one of the one of the patents that

A.N. (1:35:58)

could be my talking a little bit longer. So we could talk a bit about this shirt. Yeah. I talked about that something at the hour to run a no, no, i Yeah. i

J.R. (1:36:07)

I'll take off a little bit after that. But yes, yeah. So the the high level, one of the patents that was out there is being able to essentially have a voting system. How did the I don't remember how was put it basically democratizing decision making. So with the, the Nova nodes, for example, you know, that the robot has several different ways that it can communicate, but then how does it know which is is the most important thing to do? So for example, you say, hey, robot, you need to take this package and deliver it to this, this destination robot thing goes down that that road and says, oh, there's a tree in the way. This boat is telling me that I need to go that way. You know, what did the other boats say? And being able to then analyze, what are all the other factors going on? What is the next thing that we do? And so that was that is? I'm like, What's your

A.N. (1:36:59)

tree? But like, it sounds like, I mean, so you could have a state machine or something. But it sounds I mean, then that could be represented as behavior a lot more. You have more logic kind of built in, baked into it. Right?

J.R. (1:37:12)

Okay, a lot more. Exactly. And a lot more real time type logic, like, what is going on, right now, as the environment is changing? Because the environments changing all the time, like, do I stop? And so a better example would be like, Okay, I'm, I'm supposed to deliver this package, but I see this, this six year old girl coming towards me, do I stop and talk to the girl? You know, based on what based on? Oh, I've got time? You know, okay, yeah, that would vote. Okay, we're gonna vote, we're gonna stop for 30 seconds and say hi to this girl, versus just blowing past because I've got something to do. But sometimes it can be known, this is a higher vote, because you need to get to this destination by this period. And you know, so. So that that is a big piece of kind of the patents behind it is being able to have a voting system. Yeah. I mean, to know what to do next.

A.N. (1:37:58)

Interesting. Yeah. I mean, the voting system, like you could you could figure out an, like, several ways of scheduling, what would be important. It could be probabilistic. It could be like a integer voting system count. It could be real numbers, or whatever it might be. Yeah. And then it sounds like it's a movie. So it sounds like you have more of a behavior tree setup, where it's like, I'll choose this or this or this based on different options that are available. And that's very baked into this, this kind of conditional logic of what to do. So maybe it has a nice representation of how to encode these kinds of behaviors, or something is that so it has a nice way for programmers to define what it is to that should be considered in making decisions or how to prioritize? Yes, yeah.

J.R. (1:38:55)

Yes. I'm sure that Wilkinson is our he was a, he worked at Tesla for a while, and I'm sure he would be shaking sensing. Jason, you're totally butchering it. He was one of the main architects of it. But anyway,

A.N. (1:39:09)

so with that, I noticed right now, you don't have very much on your on the company GitHub website. Yeah. It would be very nice if this that you're talking about would be open source. So that can be poked around. And we can look at it in this kind of thing, because that would be very cool. What Yeah, what is on there is mostly JavaScript, which is interesting to me. Could you tell me a little bit about the decision? I don't know. It's like your operating system. A lot of the code that you're running is running in like a Node js framework, which is like a, but is it server side JavaScript framework?

J.R. (1:39:53)

Yeah, we get a lot. We do do a lot in Node js.

A.N. (1:39:55)

So Gotcha. Yeah, you Interesting. So it's gonna be like a no JS Ross like thing that has behavior trees. Would that be a way to think of? What do you call it again, called DAX.

J.R. (1:40:11)

So DAX is just like the robot XO

A.N. (1:40:12)

s. Okay. Yeah. Yeah. Gotcha. Cool. Yeah. Any timeline on open sourcing that? Because around No,

J.R. (1:40:22)

I don't have a timeline. And, unfortunately, right, you have to wait, you know, the business, the business decisions versus, you know, it's somewhat similar, similar with patents, you know, anybody that that has dealt with patents or, you know, litigation around that, patents aren't necessarily the end all, it's not like, hey, we invented this thing. And nobody, it's actually more of a defensive mechanism. So if a company decided that they wanted to file a lawsuit with us and say, Hey, we have this patent on this thing, and it's really a, well, we have a patent to, you know, so this is why we're allowed to operate in this space. So, so that that is, from a business standpoint, that's the same reason that patents exist. But the other piece of it too, obviously, is protecting intellectual property, I get that. But in the same way, like, you know, writing, you know, source code and things like that, if you let out too much of the secret sauce, when you're in like infancy type stages, you know, like this, it could, could put you out of business, not super likely know, that somebody would take and be like, Oh, we're gonna do exactly what DAX did. But the but the founder, that's definitely one of the things that he absolutely loves, is, you know, wanting to make this open source. So I don't know what the dollar amount is, or the the comfort level is to the point where he's gonna want to open this up to the world. But that's something that he talks about pretty regularly, you know, wanting Okay, well, that'd be really cool. It

A.N. (1:41:42)

would probably be a good way to battle hardened it to absolute like you guys are running it. And what I mean, so you said you currently have about 10 robots. By the time you have 1000 robots or something like this, it'll probably like, the software is going to get a lot more battle hardened, because you'll just have run it through so many things, but the community probably could help in that process. Oh, yeah. Absolutely. Like, it's amazing. To me some of the things like I mean, there's so I guess, I don't know what's public that I can say. But there's a lot of very, I would not expect open source to be used in a lot of really critical projects that are done by companies that have a lot of stake in reliability. And I'm very impressed that a lot of open source software is being used there. So it does battle hardened. And so it might be a benefit for that reason to and then also people would have a better point of comparison to, to look at it and see if it's useful and this kind of thing. And that would be awesome. Yeah, absolutely. Okay, so now we are going a little over time. What do you I guess, wrapping up any links or contact info or anything you'd like to share with our listeners?

J.R. (1:43:07)

Sure. Yeah, I think that, again, you know, those that are listening just via audio, you know, can't see the robot behind me. So if you go to DAX bot.com, that's D AX vo t.com. That's, you know, you can see kind will

A.N. (1:43:23)

links with the episode so that they can just click if they look in the description.

J.R. (1:43:28)

Yeah. And then those of you that that are so inclined, start engine comm slash dexpot is where we're doing our crowdfunding raise. And you can take a look at kind of the kind of the more the business plan around it, you know, kind of the the need and the focus that we're, we're filling and all the legal stuff we filed with the Securities Exchange Commission. So anyway, those would be the two things tax bot.com and and start engine slash tax bot. And yeah,

A.N. (1:43:55)

appreciate that. All right. Awesome. Been great interviewing you. Yeah, you too. Thanks so much. Audrow. Thanks for listening to this conversation with Jason Richards. Thank you again to our founding sponsor, open robotics and I hope to see you next time