PwC's Tech While You Trek: Drone Analytics

Tune into another episode of Tech While You Trek to hear PwC Director Kumar Satyam talk about how drones are helping various industries collect data in a smarter, faster way and risk-free.

Transcript

Tech While You Trek - Drone Analytics
Release Date:  June 14, 2021 

Adam  (00:08):

Hello, everyone, and welcome back to another episode of Tech While You Trek. I'm your host, Adam, and today I have with me Kumar Satyam to talk about using drones to collect data. Please introduce yourself, sir. Thank you for taking the time. Tell us a little bit about how you came to be with PwC.

Kumar Satyam (00:23):

Thanks, Adam. I have been part of PwC for eight plus years now, and I lead the analytics in PwC, a lot of work around AI machine learning and some of the emerging tech via PwC. 

Adam  (00:36):

So, when we talk about drones, the flying, remote-controlled version is immediately what comes to my mind, but obviously, the technology is much broader than that, right?

Kumar Satyam (00:44):

You are right in pointing out, Adam, that what we see is usually the flying one, which in broader sense is called unmanned aerial vehicle technology. But there are drones that can go under the water and so forth. Let's just focus on the aerial one, because that's what we all are familiar with. So the technology is pretty broad in the sense if I just think about from the perspective of how do you operate. The simplest one is what is called the manual one, where we play in our backyards with the remote control and you're manually operating. A semi-autonomous one is a little more sophisticated than that. You basically can tell it where to go and so forth, but you are still monitoring it and correcting the path of travel and so forth. And then the last one, which is not very prevalent, especially in the US, is a fully automated one where you just define everything and let the drone do what it wants to do.

Adam  (01:34):

Talk to me about some of the challenges that arise in using these drones and applying the technology to data collection.

Kumar Satyam (01:41):

The one thing that is a challenge is around regulation and how do you ensure the safety around it. If you know the aircraft and aircraft systems well enough, they're mature technology, right? So they talk to each other, and then you have heard so many times in the news two aircraft came close to each other, but then the system basically alarmed it, and actually they moved away from each other and so forth. Those type of technologies don't exist right now in I would not say not in any drone, but most of the drones. So now basically you have the issue around safety when multiple drones are flying.

Now the second most important one, which is more related to the folks like you and me, is the privacy of the data. When our drone is flying, especially at a low height, it's gathering a lot of information, right? Because, by nature of it, drone has a very strong 3D camera, so it's a very high-resolution picture of it. Now there are concerns about all the data that the drone is gathering on the way from, say, origin to destination. Who owns the right to it? If it's flying above my house, does that mean I own the right to that data? And can I ask the drone operator to delete it or the drone operator owns the right to it. And the third one is basically start using it for criminal purposes and so forth. So there's a concern about how do you prevent that?

Adam  (02:59):

As all of this activity increases, as all of these opportunities for usage are increasing, what are some of the trends that you're seeing?

Kumar Satyam (03:06):

Let's start with the one that we probably are most familiar with, which is around transportation or shipment of the goods. We know that a lot of the big transportation companies are looking into using that. The other one is very common nowadays is people using it for doing inspection and maintenance. If you have to basically check a bottom of a bridge for cracks. I mean, in the past, it used to take a lot of time for a human being to go down and figure out with a telescoping camera if there's a crack or not. Now you can fly a drone and he can take a very high resolution 3D picture and tell you basically, is it fine or is actually the cost of maintenance on this one. The third one that's been picking up is lot in the infrastructure world that we talked about construction and trying to see the status of the construction and doing a 3D imaging of the ground and so forth.

Adam (03:59):

Well, so not only are these drones creating whole new opportunities in terms of data they can collect, but it also sounds like they are being used to take humans out of potentially dangerous data collection positions.

Kumar Satyam (04:13):

I would put it differently. I would say they are enabling humans to do a better data collection without putting life at risk.

Adam (04:19):

Okay.

Kumar Satyam (04:20):

What you are able to do, you're able to do the work faster, smarter with more accuracy. And I'll give you another example, right? I mean, in the medical industry, one of the new trends in a very remote areas, the challenge used to be, to basically take the medicines or the medical equipments there when the roads are not good enough and it takes you days to do it. Now with the drones, what the companies are trying to do, a lot of dangers are trying to do is have them delivered to the place very fast. And I would use this as an example where it's saving life.

And they've got other good examples, like the crop one. I mean, people used to be subjective and looking into the crop and saying, is it actually suffering from pesticide? Do I need to spray something? And now with the state of the art 3D technologies, you can fly drones above the big farms, take good pictures and contrast and compare and say, which areas need to be addressed in, which are good. And there's some foreign areas I will give you. I mean, I didn't realize that in India, people have started using drones for doing the photography, video and photography of the weddings. That would be very interesting. I mean, if you have money, you can get it done.

Adam (05:24):

You bringing it up people. It's still cheaper than ordering a helicopter and hiring a pilot, right?

Kumar Satyam (05:28):

When you have a helicopter flying at a hundred meter it's going to scare everyone out. Right?

Adam (05:34):

And it's safer for everyone. And back to that original point. Well, so what are some other ways that you're seeing drones being used to solve business problems. Solve more outside of the box problems like this?

Kumar Satyam (05:43):

But a film industry uses it. Right? A lot of the aerial photography and videography to do. When you can fly drones above the forest for acres and acres of the land, and then basically, take them half the pictures to see if there's a chances of a fire starting somewhere or not. And if the fire is starting, then they can quickly spray the chemicals right there, which is sometimes harder for when the aircraft to go in.

Adam  (06:06):

Are you finding that a lot of people are having to be trained specifically to be able to basically monitor what a drone is flying over and to interpret what they're seeing or is it the kind of thing where that data is streamed to a bunch of eyes all at once? How does the data get transferred from what the drone is seeing to where humans begin to interpret?

Kumar Satyam (06:22):

You do have to train it, right? It's like, when we training a child to say, what is good versus what is bad? What is right versus not right? So training is a part of it. Okay. You have to train the drone that you have to train basically, a human being, how to use the data. And it's a cultural shift, right? I mean, if someone has done the work manually for 40 years, and you're asking him to use the technology, you have to train him. It's a process change that has to do. In terms of transferring the technologies, if you have large volume of data, usually multiple ways to do it. Depends upon the use and depends upon the maturity of the company that you are using. 

Adam  (06:57):

Well, so this sounds like it would be prime territory to introduce AI into the situation. Talk a little bit about how AI and drone technology can work together.

Kumar Satyam (07:07):

AI is a good enabler for it. And let me give you a simple example. Okay? If someone came in and basically hit my car, for example, right? Now the insurance company guy comes in and how would he know that the car has been hit or not hit? So what it requires, it requires you to have a whole set of what we call the perfect pictures of the not so perfect pictures. And then what the machine learning and the deep learning algorithms do, they actually “create” what is called the discriminating factors or the anomaly factors to figure out how do I actually figure out if something is not right?

Adam  (07:42):

So you would input all of these pictures of the bumper of a 2010 X brand car into this basically digital database. And then it would take pictures of the supposedly damaged bumper in question and compare it to its database of what it knows to be perfect bumpers and identify the anomaly?

Kumar Satyam (08:02):

You've got it right. I mean, essentially it does that. You basically have put the 2010 bumpers, right? The perfect one and not so perfect one. It starts looking at it. And it says, basically, I see a difference here. This doesn't seem right. Doesn't fall into the pattern that I'm seeing. In fact, I mean, that's how basically, a lot of the utility companies looks into the post that we talk about, the transformers they talk about basically, because what they do is train by using the older pictures they have where are the folks had taken it. And then basically as the drone starts flying, I mean, it's learning as well, right? It's a machine learning. It's learning as it's flying. It's taking pictures as well. Basically starts learning basically from that scene. This picture matches to a good picture from the past and this picture doesn't match. And hence it will flag that. The beauty here is of machine learning uses as it actually does more and more, it learns more and more, right? Because it’s capturing a much higher resolution picture and a video that can be useful self-training going forward as well.

Adam (09:03):

So shifting the gaze forward, how do you see drone technology converging with other technologies?

Kumar Satyam (09:10):

You will see basically merging with a lot of the aircraft technologies that is there. I mean, if you're familiar with it, it's called, a system called TCAS is what the aircraft uses. Right? At some point I think a larger version of the drone would be a flying car. A lot of the imaginary stuff we talked about, real-time data processing, like sensor data that we talked about and other areas you would almost have a, real-time saying of a drone is flying and say on the other coast, you will be able to get a real time feed and information for it. And the drone would be able to make a decision of almost like a self-driving car. It can make a decision saying, “I think there's going to be a fire here. Let me go and spray the chemicals right away.” So a lot of the autonomous stuff that we hear about in driving cars and those stuff start happening.

Adam  (10:02):

So Satyam before I let you get out of here. I like to ask all my guests one fun question. Are you ready?

Kumar Satyam (10:09):

Yeah.

Adam  (10:10):

What would the you, of 20 years ago be most surprised that you use technology for in your modern daily life?

Kumar Satyam (10:17):

I would say, the digitization of stuff. One example I will give you when I did my undergrad, when I was doing my thesis, I had to go to library and print out the journal papers and everything. Right? And do it. Now I mean, if I look 20 years down the line, if I really want to read a paper, what do I do? I just open up a laptop, I go and just do it. Another fun example, which I didn't realize until my son did it is the whole touchscreen stuff.

I mean, there's no concept of analog. I look at like 1987 and so forth. My son, he was two years old. And I mean, we were watching television and he starts going to the screen and starts poking the screen because what does he think? He thinks has been to be a touchscreen, which is basically you can change the channel. So I cannot imagine that 20 years back, I mean, I showed him the analog phone and he was like, what is this? I cannot imagine the life 20 years back. I don't know. I mean, if we go back, how would we feel? But that's the one thing, I mean, I'm really surprised by.

Adam (11:12):

That's a fantastic answer. Well, listen, thank you again so much for your time. Thanks for stopping by. We appreciate it. 

Kumar Satyam (11:17):
Thank you. 

Adam (11:18)

This has been another episode of Tech While You Trek. I've been your host, Adam, and we'll talk to you again next time.

Speaker 3 (11:27):

This podcast is brought to you by PWC. All rights reserved PWC refers to the U.S member firm or one of its subsidiaries or affiliates, and may sometimes refer to the PWC network. Each member firm is a separate legal entity. Please see www.pwc.com/structure for further details. This podcast is for general information purposes only and should not be used as a substitute for consultation with professional advisors.