Retrofitting Equipment for Remote Control: An Interview with Teleo’s CEO

Bryan Furnace (00:00):

Today, we're here to talk about AI technology again because, man, it is coming. It's coming fast. And here to talk with us today is Vinay Shet, from Teleo, which is another company that is able to retrofit existing equipment so that it becomes autonomous. And I don't know anything about this stuff, so I'm going to just flip it over to the interview and let the expert talk about this.

Vinay Shet (00:31):

Teleo builds technology that lets us retrofit our customers' heavy equipment, like wheel loaders, dozers, trucks, that allows us to first remotely control those machines. That allows an operator to sit at a remote control desk and operate the machine remotely. But in some situations, some use cases, to let the machine run autonomously by itself and only jumping in when the machine needs help, or when they're doing a specific type of work that requires more finesse.

Bryan Furnace (00:57):

And so at this point in time, are we talking primarily, hauling-type of equipment that are being equipped with this, or is this something that we could throw on an excavator and an excavator could dig autonomously?

Vinay Shet (01:09):

So, the technology itself that we've developed is machine agnostic and OEM-agnostic, so it doesn't really matter whether it's a dozer, or a loader, or an integrated dump truck. It doesn't really matter what OEM manufactures the truck, whether it's a Caterpillar, John Deere, World One, and so on. We are focused a lot, however, on the use case. We like to see use cases which are repetitive in nature. So we specifically focus on dirt moving jobs, for example, in site development for construction or in quarries, where the same actions are done by the operators over and over again.

Bryan Furnace (01:40):

So, just thinking about this from the standpoint of a contractor, what kind of time is my equipment going to be down for, for installing the system and getting it all set up? How long is that going to be out of commission before it's ready to rock?

Vinay Shet (01:53):

It's actually been very little time from the customer's perspective. We tend to do a lot of our retrofitting work after hours or on the weekends, and that allows us to make sure that our customer doesn't really see downtime, and that we are able to bring the solution to their machines as quickly as possible.

Bryan Furnace (02:06):

Oh, so interesting. So you guys could be in the middle of an install, and I can still go out tomorrow and use that machine in the middle of the install. You guys kind of button up for the day, if you will. I get to go run it and then in the evening, I just leave it where it's accessible. You guys can go back in and finish?

Vinay Shet (02:20):

Exactly. So we do our install in phases, where we put certain parts of the gear on, but it allows the contractor to run the machine during the times in between. And what's also interesting about this solution is that even after a kit is on the machine, they customer, if they want to drive the machine in manual mode where they are in the cab, they can do so. So our installation is very clean. It is out of the way. And so, the operator can still sit in the cab, they hit an e-stop button, and they're able to take on manual control of the machine if need be.

Bryan Furnace (02:48):

Now you mentioned that there is for the autonomous machines, we've taken the operator out of the cab, there is still someone overseeing that machine.

Vinay Shet (02:57):

Yes.

Bryan Furnace (02:57):

Can you speak more to that role, as far as what are we looking at for training time? How many machines is one person potentially going to be responsible for? That kind of thing.

Vinay Shet (03:08):

Yeah, so I think it's very important to stress that what Teleo has built, is what we are calling supervised autonomy. What this means is that the human element, the human operator element is a critical component of what we have built. And so, the expert operators today would be able to guide the machine remotely, and be able to tell the machine where it needs to go and control this trajectory, and so on. But only in some situations where the operation is highly repetitive, that they can then put the machine in autonomous mode and basically supervise it only. And when it is running in autonomous mode, they could then go and control a second machine. In doing so, now a single person can then manage multiple machines at the same time.

Bryan Furnace (03:47):

So you could potentially have a fleet of 15 or 20 trucks, and there's only two or three people monitoring all of those trucks.

Vinay Shet (03:55):

Yeah. So we like to see it as the number of operators operating a certain number of trucks. And in some situations, it could be one person controlling one truck, but in some of those it could be one person doing two or three. It really depends on the use case, that the customer is deploying the technology in.

Bryan Furnace (04:13):

Now how complicated is this? I think about getting involved in an AI technology situation, and this is going to be months of training. But for me, a rookie that doesn't have any experience with AI stuff, but I am an experienced operator, how long would it take to train me up to where I could use this equipment?

Vinay Shet (05:45):

What we have done is we have tried to absorb the complexity of building the solution on our side, on Teleo's side, so that it becomes dead simple for the customer and for the operator. What we have seen is, especially with the long range remote control aspect of it, an experienced operator is able to get behind our controls and operate the machine pretty much immediately. The controls are similar. The machine behaves exactly the same way. They can do with the machine remotely what they can do sitting in the cab itself, and therefore it is completely familiar and they're able to be deployed almost immediately. On smaller machines, we have seen experienced operators be productive, literally in minutes. On bigger machines, we like to get them used to it over a course of a few days, and then in under a week, they tend to go at capacity, at speed.

Bryan Furnace (06:29):

Oh, wow. So, you're talking a week or less, you can potentially be trained as an operator to run these things?

Vinay Shet (06:34):

Yes.

Bryan Furnace (06:35):

So that does sound like you guys have absorbed the majority of the blow, when it comes to getting the machine set up, if you will, on a job site. And so my next question is, and it's really a two-part question, there's two scenarios in my mind. The first scenario is, we're on a job site where we've been hauling material out of the same stockpile for the last three days. But now that stockpile is done and we're moving to another side of the job, and we've moved to a different fill location. Is that going to require me calling you guys to come out and reprogram the route for the trucks to take? Or, is that something that I can do as an operator?

Vinay Shet (07:08):

So the sites that we operate in, and in addition to the technology we deploy on the machines, we also deploy a mesh network that allows us to communicate with the trucks and with the equipment. So to answer your question, if the customer wanted to move to a different location of the site, they would need to make sure that the network has moved that location, or they've deployed a network widely enough that the entire site is covered. But if they want to operate in a different location, because it's a supervised approach, the operator can drive the machine to that location and the system can then function there.

Bryan Furnace (07:39):

Now, this is the second part of the question. Now, we're done at this job site entirely. We're moving 10 miles down the road, reestablishing an entirely new work site. Is that something ... In my mind, that's probably where I'm going to call you guys and say, "Hey, we need to switch all the gear over, get everything primarily set up so that our operators can take over." What's the process there?

Vinay Shet (07:58):

So we actually are working closely with a very strong dealer network, who would be involved closely with us and the customer, to exactly support situations like this. Where if they want to redeploy the fleet somewhere else, then they would work with us and the customer to make sure that that happens smoothly, so there's no downtime for the customer.

Bryan Furnace (08:16):

So this is where, my next couple of questions, I ask you to look into your crystal ball, into the future and speculate. And I will say that for the audience, this is entirely speculation, but being on your side of this industry, what does the future look like for AI really rolling out on the civil construction side? Because right now, it seems like a lot of this is focused on kind of mining inquiries, because there's a lot of repetition. But really, the majority of our audiences is regular dirt contractors. When do you think we'll start to see this technology really start to gain a foothold on the civil side?

Vinay Shet (08:53):

I think, we are at the very beginning of that process, and I foresee that over the next several years, you're going to see more and more use cases being tackled by technology like this. We are already looking at samples, site development types of projects, but then if you transition that into large construction work, and this could start to apply there pretty rapidly as well. So, I think it's coming pretty soon.

Bryan Furnace (09:14):

I would almost argue that it's here now. I mean, you're already starting to roll out on some of these site development projects that you mentioned. I think it's going to be one of those things that just kind of sneaks up on us and we turn around and go, "Oh, that's a truck without a driver in it." So, my next question, and this is my final question, and this is pure speculation.

Vinay Shet (09:36):

Sure.

Bryan Furnace (09:37):

Is everyone in our industry going to lose their job because robots take over? That's the giant fear of everyone in our industry.

Vinay Shet (09:44):

Absolutely not, right? In fact, very specifically, the way we are approaching this is the human operator supervision aspect of it. We expect that we will need more people in the future than we need now. Today we see across the board, that there's more work to be done than there are people available to do that work. And really, we see this technology as an enabler to take those folks and let them actually do more. Even if you ignore the autonomous aspects of it, we enable a world where operators can sit behind a control station and operate machine. That step alone, the transition step alone, we think can help attract more operators into this world, into this industry.

(10:20):

We can imagine a world where we attract more women operators. Imagine a world where even operators who may have hurt themselves, who may have disabilities of some kind that prevents them from getting onto a machine, can now become productive in this industry now, because they're able to sit behind our controls and run those machines. And so I think, this opens up many, many opportunities. And I couldn't be more optimistic about the future of operators working in this industry, using this technology as a leverage to do even more than they were doing before.

Bryan Furnace (10:48):

And that's been the hardest thing to communicate, because that is exactly how I foresee this rolling out is, it is something that is assisting the industry do something that they're already doing. It's not, everyone's going to be out of a job and these machines are going to come in and take over. Roles may change, and the way you interface with the machines may change, but at the end of the day, the work is still there to be done and people have to be there to do it. You can never create an AI robot that will be able to make those sort of intelligent decisions on the fly.

Vinay Shet (11:16):

That's right. That's right.

Bryan Furnace (11:16):

Well, Vinay, thank you so much. This has been really informative and it's been a pretty unique perspective seeing someone on your side of the table that's actually developing this stuff, talking about it.

Vinay Shet (11:25):

Yeah, it's been exciting.

Bryan Furnace (11:26):

There you have it. AI technology is working its way onto our job sites. But as you can see, it's not the big, bad, scary monster that a lot of people think it is in the industry. This is going to be something that is here to shore up the fact that we are suffering from a huge worker shortage. If we can make our manpower go further, we're going to get more work done and we're not going to be hurting so bad for people. So this is ultimately going to be a good thing for the industry.

(11:51):

So as always, I hope this has been helpful. Thanks for watching, and we'll catch you on the next episode of The Dirt.