Fact vs. Fiction – The Truth About Autonomous Equipment

00:00:00:09 - 00:00:26:12

Bryan Furnace

Today we're here to talk about autonomous equipment again. It's been a while since we've brought this subject up. Today we're talking with SafeAI just to give us a better understanding of what's on the horizon and how it's going to impact the workplace.

00:00:26:14 - 00:00:31:04

Bryan Furnace

Brought to you. Thank you so much for taking time to be on the dirt today. I really appreciate your time.

00:00:31:06 - 00:00:39:06

Bibhrajit Halder

Thank you, Bryan, and really excited to talk to you. I saw some of your previous episode. You are doing an amazing job, so looking forward to talking to you.

00:00:39:07 - 00:00:50:23

Bryan Furnace

Thank you. I really appreciate that. So for the audience, can you give us just kind of a brief overview of what SafeAI is doing in the excavating industry and kind of what you guys do overall?

00:00:51:01 - 00:01:14:22

Bibhrajit Halder

Absolutely. So we started here about seven years ago, about 100 people based in Santa Clara, California. What we are doing, we are going after the mining, construction, kind of the heavy industry, fuel taking that existing machine, truck, dozer loader, retrofitting them to make it fully autonomous. So it is a fully autonomous, really taking human driver off the unsafe area.

00:01:15:00 - 00:01:28:07

Bibhrajit Halder

No human oversight, no human in the loop whatsoever, making it fully autonomous, giving back to the operator whether it is a contractor or a subcontractor so they can get the job done more safely and more productively.

00:01:28:09 - 00:01:38:22

Bryan Furnace

And so I guess the big question I have, because this is such new technology, is what kind of capabilities does this equipment have at this point in time?

00:01:39:00 - 00:01:58:14

Bibhrajit Halder

Yeah. Then we, you know, think about it maybe a little bit about that. Tesla might not be a good example, but let me use Tesla's example. Like passenger vehicle guys already developed autonomy at a very high quality even if it is not fully there. But their problem is so much harder. They spend about $100 billion developing this solution.

00:01:58:16 - 00:02:20:18

Bibhrajit Halder

So why started from there? We are actually the bridge between what's going on in passenger bringing back to the heavy industry. So on the three critical piece, you need to drive things autonomously. You have a high quality of perception. You know what's going on around you. So you have a very high quality of perception. You have camera, lidar, radar, and then you have a high quality of decision-making process.

00:02:20:19 - 00:02:37:06

Bibhrajit Halder

What should we do? So all that factored into the solution, and the third one is all the component we had has to be very cost effective. We cannot put a, you know, half $1 million worth of hardware and a half $1 million worth of truck. Sure. So I think think about all those come. So these machines are very smart.

00:02:37:08 - 00:02:43:09

Bibhrajit Halder

They can do everything by themselves. Now we are picking use case by use case where this should be deployed.

00:02:43:11 - 00:03:00:07

Bryan Furnace

Gotcha. That's kind of one of the other thoughts that I had is with this being such a new technology, it seems like as a layman, it would be cost prohibitive. Are you finding that's not really the case? Now that we are getting to the era where where this technology is finally starting to be implemented?

00:03:00:12 - 00:03:09:14

Bibhrajit Halder

Yeah. Let me set the stage up a little bit. So back in 1990, the autonomy was already being used in the by the US defense.

00:03:09:16 - 00:03:10:10

Bryan Furnace

Okay.

00:03:10:12 - 00:03:35:13

Bibhrajit Halder

So like a lot of the technology we use today like internet, your GPS is same story, right? So US military is already using autonomy back in the 90s early 2000. They have something called DARPA's Grand Challenge. Now like this is amazing. Can we also give it to this other people who does things like driving, truck driving, tractor driving car and that is really the spark, that whole thing.

00:03:35:18 - 00:03:36:16

Bryan Furnace

Interesting.

00:03:36:16 - 00:03:41:17

Bibhrajit Halder

So why I started from there. This has been in the making for almost 25 years.

00:03:41:18 - 00:03:42:15

Bryan Furnace

Wow.

00:03:42:17 - 00:04:03:16

Bibhrajit Halder

And was part of a company called caterpillar. Obviously, you know, being this industry, I was there with the eight in a seven years. So Caterpillars of the world, Caterpillar, Komatsu, Hitachis of the world. They have been working on it for 20 years. They delivered the first solution in the market full production ready almost ten years ago.

00:04:03:18 - 00:04:05:07

Bryan Furnace

Wow.

00:04:05:09 - 00:04:36:02

Bibhrajit Halder

And that industry has driven I moved 10 billion tons of material fully autonomously. No human in the loop at all. They’ve given 20% productivity improvement, 25% operating cost reduction and zero safety incident for almost ten years. But at a very high price point. So yeah, so that was the one almost unaffordable to most of the construction industry because they cannot pay that price.

00:04:36:04 - 00:04:53:15

Bibhrajit Halder

And the reason I'm making a long-winded answer, because this is going to set the stage up the way you want to think about that solution is like a I give you example like that's your Tesla Roadster. Amazing car. Okay. But not very many people can really afford a $200,000 car. Sure. Now you think about what we are bringing.

00:04:53:15 - 00:05:13:00

Bibhrajit Halder

What? Well, autonomy today for heavy industries model three interesting. Model three is that the model three is better than Roadster but much cheaper and also better technology. So to your point is the price is affordable today for this industry because we have been working on it for the industry for 25 plus years.

00:05:13:02 - 00:05:28:05

Bryan Furnace

So if I'm translating this correctly, the really the big reason we haven't been hearing about this on the front lines until recently has nothing to do with where we're at in the development of the technology. It's more it just hasn't been practical from a cost standpoint until now.

00:05:28:06 - 00:05:52:06

Bibhrajit Halder

That's right. And I'll give you another punchline on that. Like I say, if Google was doing autonomous construction in a heavy industry, we would have had 30% of the fleet autonomous by now. So two things, right. The cost that there is not enough interest went into that. And as I mentioned, the passengers paid 100 billion, got invested in that ecosystem jointly.

00:05:52:08 - 00:06:05:10

Bibhrajit Halder

How much you think got invested into our vertical, which is heavy industry and construction and all that is about 2 to $3 billion. Yeah, yeah. So one 800 billion versus 2 to $3 billion. That's another reason. Yeah.

00:06:05:13 - 00:06:28:20

Bryan Furnace

Interesting. So it's really the return hasn't been there up until this point to really make it worth pursuing on a large scale. That's right. Interesting. I never would have guessed that. So I guess now knowing that this technology has actually been around for a while and it's far more advanced than I ever anticipated it being. What sort of contractors would really benefit from autonomous equipment?

00:06:28:20 - 00:06:31:22

Bryan Furnace

What kind of industry should really be looking at this equipment?

00:06:32:00 - 00:06:56:09

Bibhrajit Halder

Let's think about the construction industry. Let's figure one example of you're building a dam, right? Whether you're building the GC as a subcontractor, it's a series of people. The way you want to look at broadly is that a typical dam, let's say, takes about five years to build $100 million to, you know, cost. If you deploy autonomy into this system, you're going to get the job done in 40, go in about $75 million.

00:06:56:11 - 00:07:15:08

Bibhrajit Halder

So you are reducing 20% of the time and about 20 to 25% of the cost. So as a contractor, as a part of that industry, you're looking at autonomy is going to allow us to get the job done faster, cheaper, and I can do more. You know, as a subcontractor, you say I can only take on 100 project a year.

00:07:15:10 - 00:07:41:10

Bibhrajit Halder

That's my human capacity of the company capacity. Now all of a sudden you can take 150 project a year because now you have added advantage. So really it is doing more for less and you will see the world is happening. I can give you another example. Just a way you want to look at it broadly is that 20% of the world is developed well, whether it is building bridges or not, 80% of the world is looking at the 20% of the world.

00:07:41:10 - 00:07:48:21

Bibhrajit Halder

Say, I want the same thing, but instead of taking 30 years, I want it in ten years. And that is what this autonomy is going to enable.

00:07:48:23 - 00:08:08:20

Bryan Furnace

So when you're talking about the I understand the the cost is coming down because of the human element, but is it also getting completed in a shorter time because these autonomous machines can run around the clock? You're not dedicated to working hours for people, or what's causing that to be more efficient, that you're able to get the work done quicker?

00:08:08:22 - 00:08:33:08

Bibhrajit Halder

Yeah. Three main component and I'll give you on the order of magnitude impact on that. Right. The number one is yes. It runs longer. Our right. Okay. If you take a typical a construction project and we are in Japan, we are deploying. One of the idea is that they run about nine hour shift a day, but they have permits for 12 hour, but they don't run the next three hour because they don't have people that nine-hour shift.

00:08:33:10 - 00:08:51:03

Bibhrajit Halder

Right. They can't hire people for just three hours, and they do over time sometimes when you need it. Now all of a sudden, we are going to deploy autonomy for last five-hour shift. So it lifts to think about a nine-hour day. We are going to add last five hour. So take two plus three and we are going to add additional revenue to that.

00:08:51:05 - 00:09:08:04

Bibhrajit Halder

That just to getting to it because we don't want to immediately switch it up. That's a practical way of getting into the say. All of a sudden, we are giving them extra 10 to 15% a day productivity. And you can imagine what it does to the subcontractor who is running on a very tight budget. I need to deliver XYZ.

00:09:08:06 - 00:09:31:02

Bibhrajit Halder

So that's why you will see they are going to give away anywhere from 2 to 10 in a 2 to 10 hour, depending on the where you are working extra productive. Our second is autonomy. The way I say autonomous machine. Another good thing about autonomous machine it does what it's supposed to do compared to a human. So what I mean by that if I am driving a truck autonomously, the truck is exactly going to do what it going to do.

00:09:31:07 - 00:09:38:20

Bibhrajit Halder

Now, my loader and dozer are very coordinated. So I am actually gaining about 15% productivity on better coordination.

00:09:38:22 - 00:09:47:08

Bryan Furnace

And your truck doesn't randomly drive over to the convenience store to grab a can, a monster and a Snickers bar. This is going to always be there doing exactly what it's supposed.

00:09:47:10 - 00:10:08:16

Bibhrajit Halder

As you know, and I can I being the probably more mine than I should have more construction. It should have last 20 years. I can tell you all kinds of story. What what they do when they're in that cab. No. Yes, I'm sure I will have I can tell you show us all of them. I cannot tell you in your show, but that is a lot goes on in the cab, in that truck.

00:10:08:18 - 00:10:27:19

Bibhrajit Halder

Now, I don't mean not to any of that, any of that. And the third component you get advantage. Is that because it runs better. And I'll tell you what I mean by runs better. One example like in typical man one truck, they do a lot of slight slipping and they intentionally do slides something because the speed up slides slip and get get the job done.

00:10:27:21 - 00:10:49:11

Bibhrajit Halder

That takes a toll on that. Just physically on the tire, on that chassis of the body, it takes about 10% of the maintenance cost goes up. Oh wow. For us, autonomy will not do the slide slipping. Autonomy will do what it's supposed to do. So we actually benefited about 15 to 20% or sometimes 25% of the maintenance cost goes down.

00:10:49:13 - 00:10:56:01

Bryan Furnace

Interesting. So you're really getting kind of multiple prongs of savings there.

00:10:56:03 - 00:11:14:20

Bibhrajit Halder

And just to kind of where I'm getting all the statistics, as I mentioned, because this industry or a lot of people don't have driven, as I mentioned, 10 billion tons of material. They moved, they drove 24 seven for the last ten years. So these are actually actual statistics from the site. Most of the construction could not get that because too expensive.

00:11:14:20 - 00:11:16:11

Bibhrajit Halder

Now they are going about to get that.

00:11:16:13 - 00:11:37:15

Bryan Furnace

What does the implementation of one of these pieces of equipment look like? I would assume that you kind of to an extent, you're restricted by the job site conditions because there's a fair amount of programing that's got to go into this and it needs to follow a specific routine. It's not something like a road crew that randomly decides they're going to move two miles down the road and work on a different area of the job.

00:11:37:17 - 00:11:57:09

Bryan Furnace

It's not like you can just move down there and and push the go button. I'm assuming there's a fair amount of setup that needs to take place.

00:13:34:19 - 00:13:53:22

Bibhrajit Halder

There is a fair amount of setup. I think when we look at a certain use case or a certain job to take on, we have certain characteristics we look at. So the number one is that it has to be a longer project. So if it is and I say project runs for six months, not beneficial because you need a little longer time because you do have a setup time.

00:13:54:00 - 00:14:13:12

Bibhrajit Halder

Second thing is that there has to be enough repeated tasks going on. If you are kind of variability all the time, human does that better when you're like literally like I went to one of the job site, they want to build a pool on your backyard. It's a big pool. And for the hotel, that kind of job is still going to be by the human.

00:14:13:14 - 00:14:34:17

Bibhrajit Halder

It's very complicated. And I went through this, you know, this guy done that for 30 years, and he was teaching me for almost two hour how they do it. You're not going to do autonomy for the pool. Designing today is just too complicated. You are looking at I am building a dam. I am building this road. I all I am doing for first let's say one year and I'm building up the dirt, dumping it.

00:14:34:22 - 00:14:56:14

Bibhrajit Halder

Flatten it out and keep doing it and doing it mile after mile. Yeah, that's the kind of job we go after because you really want to think about the passenger vehicle. They already do much more complicated stuff. It's just their problem is even harder. So there is no bar for technology in this industry. It's a question of what is the right place to go for, you know, money return.

00:14:56:16 - 00:15:22:05

Bryan Furnace

Yeah. Interesting. So that's actually a perfect segue into my next question. What does the future of autonomous equipment look like? And I kind of intended that to be an open-ended question. But now I kind of want to steer it a little bit towards do you foresee implementation of this technology getting easier to where we will see it on more of our just kind of run of the mill road jobs that are in every town, everywhere?

00:15:22:06 - 00:15:39:10

Bibhrajit Halder

You will see, I think the way I think, even in our vision, we laid out about 20 to 30% of the asset. When I say asset means any of those, those are loaded. You will see them autonomous over the next 10 to 15 years. You know, it just takes a natural time to grow. I'm not saying it will happen overnight.

00:15:39:10 - 00:15:57:15

Bibhrajit Halder

Right. This is, we have existing asset. We have a lot of things to grow. But you will see the transition happening. So you will see this more and more. And over the time and as this autonomy get deployed, people will see the ROI. The more ROI they see, the more R&D budget they can put in for future harder case.

00:15:57:15 - 00:16:16:11

Bibhrajit Halder

So it's really a cycle because you want to go into a use case that is simple. You get your ROI. Once you get your ROI, your management say, okay, I see 10% improvement. Is it 20% improvement in my bottom line? Like actual dollar, I can spend 2% of that money back into the hardware use case that can need more R&D to deliver that.

00:16:16:15 - 00:16:32:16

Bibhrajit Halder

So you'll see the cycle start. And we are going to autonomy going to start eating up more and more of the use cases. Another side of it is that doesn't mean that we are just going to do the same amount of work and less people what we are going to do, we're going to do more work in that time, if that makes sense.

00:16:32:18 - 00:16:50:19

Bibhrajit Halder

And human is going to move from this very unsafe work being on the side. And I can give you example after example on fatal accident. We see in this vertical they will move into more higher quality job where they are going to actually do the management orchestrating and planning. And that is a new generation wants to do so.

00:16:50:19 - 00:17:04:03

Bibhrajit Halder

It's actually a move good direction. You'll attract the new generation into this industry. Now you are providing them a higher quality, much higher paying and safer job then the last generation has done. Yeah.

00:17:04:05 - 00:17:21:19

Bryan Furnace

and that kind of plays into my final question, which is that is one of the biggest, I think, knee jerk reactions of our industry when you start talking about autonomous equipment is is the classic they're going to take our jobs. So that's that's everyone's defense. And so kind of as a rebuttal to that, you've kind of tapped into that.

00:17:21:19 - 00:17:31:22

Bryan Furnace

But but what else would you say kind of along those lines to kind of calm everyone's fears that, you know, overnight, we're not going to see a bunch of autonomous dozers out there doing everybody's job.

00:17:32:00 - 00:17:48:23

Bibhrajit Halder

And I think one thing I love about your work and what you have done, great show. Like you're shining the light on that thing. This industry is big. So one of the thing is that there is actually 3 to 4 people out of ten people get deployed by this, this larger construction, mining like heavy industry. So it's a big portion of the job.

00:17:49:01 - 00:18:03:19

Bibhrajit Halder

Yeah. So that's the first kind of picking up the statistics. So another thing is that we are not we will not take away a job in a sense. As I said, we will do more. And also the new generation doesn't want to do this job. You know, I go sit to the driver. I'm an operator in so many places.

00:18:03:19 - 00:18:23:00

Bibhrajit Halder

I go to the universities. The new generation doesn't want to do this job, but they are going to get interested in this thing because it's going to be technological savvy, is going to be higher paying, and it's going to be safe. Then they want to do it right. I was talking to one of the operator and say, it's a woman in Australia, and she's like, I'm just going to do it for five years.

00:18:23:00 - 00:18:41:17

Bibhrajit Halder

Then I'm not going to do this job. I'm young. I can do it now. After that, I don't want to do this job. I like I said, but if this is happening, then you get to go to office and then you're like, oh, then I am interested. So what you will see, you will actually attract more. The next generation, the current generation, there will be a displacement, but it will be very slow.

00:18:41:19 - 00:18:50:03

Bibhrajit Halder

But it's the next generation is going to get interrupted. So you'll see more people actually coming into this industry and continue that flow. It's just we are going to do more of it.

00:18:50:05 - 00:19:08:17

Bryan Furnace

It's interesting to hear you have that conversation on your side of the table, because what we've found in the industry is the exact thing you said, but it's not couched in any of those words. So what we see on our side of the industry is no one of the next generation wants to work, and every one of the contractors are out complaining about that.

00:19:08:23 - 00:19:26:08

Bryan Furnace

And then we also have this, you know, the lion's share of the people who are there going to take my job are about ready to retire. The average age in our industry is around 57, if I remember right. And so you have all these people that are complaining that their jobs are going to get taken, but they're all going to retire in the next 5 to 10 years.

00:19:26:10 - 00:19:47:17

Bryan Furnace

And this next generation that you just talked about, who's who's truly in the process of having their jobs potentially taken happen to be the generation that don't want to go out there and do those jobs, and they are better suited and to be honest, they're better equipped to take over those roles of being the programmers, being the project managers, being the technology facing portion of the workforce.

00:19:47:18 - 00:20:05:07

Bibhrajit Halder

And I think the way you set up, that is exactly right. And even for the generation that is working today, as you say, 5062 and the mining is about, you know, 52 to 52 to, let's say, 50-to-60-hour average. Their job is also not going because this technology is hard. It is not going to overtake like it takes a lot of time.

00:20:05:07 - 00:20:27:10

Bibhrajit Halder

You have hard technology, you got actual hardware, you got operation. It will take its own time to come in. So even current generation would be fine because this is not overnight technology. This is just not right. Safety is going to slow things down just by nature of it. Another way to not only just the next generation, but the the way we do mining, the way we do construction for the last 100 years hasn't been changed.

00:20:27:12 - 00:20:44:10

Bibhrajit Halder

Yeah, right. This is not the best way to build thing. This is not the best way to do things. So not only you want your next generation to actually run this thing, they have the idea that me doesn't have that idea. The one of the thing is that that's 16-year-old kid who is going to the high school, going to this college right now.

00:20:44:12 - 00:21:05:23

Bibhrajit Halder

That guy, that person has no idea how to do construction differently. You want those people to come in. So I think that some of that is not just autonomy, but is it the way we do construction? And as I keep saying, people that I don't have idea and I'm not supposed to be the guy who's coming. It's a 16-year-old kid, 12-year-old kid, 11-year-old kid now doing their thing in the high school and the school.

00:21:06:01 - 00:21:16:05

Bibhrajit Halder

They are going to come. You want to attract them, you want to bring them to the industry and the technology and higher safety is the way to bring them, will make this industry so much better.

00:21:16:07 - 00:21:36:06

Bryan Furnace

Yeah, it's I've gotten so excited, you know, having the opportunity to do the show, I get to talk to a huge smattering of people across all areas of the industry. And I always get excited talking to the tech people, because if you really think about it, our industry hasn't seen a technological revolution since we went from cable driven machines to hydraulic machines.

00:21:36:08 - 00:21:53:04

Bryan Furnace

I mean, that was a long time ago. And what we've done is we've made that technology better and better and better, but it's still the base technology of these hydraulic machines that are man operated, going out and doing the same work every day. And we've had a GPS and that's been a huge step forward. But at the same time, we're still using that base technology.

00:21:53:04 - 00:22:03:10

Bryan Furnace

And I really feel we are on the cusp of the next revolution and total evolution of the construction industry, specifically the dirt industry.

00:22:03:12 - 00:22:21:21

Bibhrajit Halder

We we will and I think this is a very good as you say, it's a right time. And another thing people do not understand, like think about your, you know, smartphone radios, iPhone and whatnot. It has gone through such a drastic change the last 30 years. You know what you used to use years ago and you're like, you use your iPhone all the time, all the time.

00:22:21:23 - 00:22:42:08

Bibhrajit Halder

But I say like I was like, do you realize what else do you use all the time? Your house, your road. Like, do you not use it a thousand a road all the time. They need to change the way we design, the way we build, the way we maintain. Because if people forget about those obvious things, like guys like you live in your house all that, what do you think your biggest purchases is?

00:22:42:08 - 00:23:00:02

Bibhrajit Halder

Your house, not your iPhone. Right. And what do you think it feels of, you know, even more than your car and your iPhones and auto? Are you still your best thing is like, I have a house, this is my thing. So I think you have a human point of view. You want to see those innovating, but how is your road and your infrastructure that you're using?

00:23:00:08 - 00:23:13:09

Bibhrajit Halder

And that's a big part of your life, not just your iPhone is your iPhone is great, I use that, but that's not the only one in your part of your life. This thing, you take it for granted. They need complete kind of new innovation and technology will bring that.

00:23:13:11 - 00:23:21:02

Bryan Furnace

Absolutely. I couldn't agree more. Well, thank you so much for the time and all of the conversation. This has been really interesting. I have really enjoyed it.

00:23:21:04 - 00:23:32:03

Bibhrajit Halder

I really enjoyed talking to you about and and as I mentioned, I keep seeing your. So keep doing it. I know this is people don't realize when they just click on a YouTube. I don't like it, like how much hard work is on the back of it.

00:23:32:08 - 00:23:39:05

Bryan Furnace

So I can't take all the credit. There's a whole team behind me. I just sit here and ask questions.

00:23:39:07 - 00:23:46:18

Bibhrajit Halder

And the guy who said that, I know that's the guy that's the most amount of work. All this on is really, really, really appreciate your hard work.

00:23:46:19 - 00:24:02:11

Bryan Furnace

Well, thank you again for SafeAI taking the time to come on the show today. And just give us a breakdown of what to expect and, and just kind of what's out there. This is a new emerging technology and there's a lot of promise on the horizon. But there's also a lot of uncertainty. So as always, I hope this helps you in your business.

00:24:02:17 - 00:24:04:10

Bryan Furnace

We'll catch you on the next episode of The Dirt.