Drones as industry disruptors

In an industry with rapidly evolving technology, what role can drones play? DNV GL talks to Pulkit Jaiswal from SwarmX about the future of autonomous drones.

Podcast: Drones industry disruptors

Referencing areas like infrastructure and asset management, Pulkit explores how drones are disrupting the way we think about inspections. We talk about the importance of real time data acquisition and how high speed data can facilitate decision-making. Importantly for the renewable energy industry, Pulkit explains how drones can help to improve project efficiency and reduce costs.


Read the transcription here


NARRATOR            Welcome to the DNV GL Talks Energy podcast series. Electrification, rise of renewables and new technologies supported by more data and IT systems, are transforming the power system. Join us each week as we discuss these changes with guests from around the industry.

MATHIAS STECK    Welcome to the DNV GL Podcast series. Today I am at the Singapore start up SwarmX and I’m sitting here with CEO and Founder Pulkit. Good afternoon Pulkit.

PULKIT JAISWAL    Afternoon, Mathias, thanks for having me.

MATHIAS STECK    Well, I’m very excited about this today, here being at your new site. We want to talk a bit about drones today, especially autonomous drones. But before we get there it would be great if you could introduce yourself, as well as SwarmX.


PULKIT JAISWAL My name is Pulkit and I’m the founder and CEO of SwarmX; and SwarmX is focused on providing fully autonomous drone technology for a vast spectrum of applications, ranging from security and surveillance, to infrastructure and asset monitor. And a little bit about myself – my background is mostly in artificial intelligence and machine learning, and those are the two key pieces of robotic systems today. And my personal ambition in life is to automate a lot of things that are really inefficient in various processes using robots. And SwarmX is really my first step to do that.

MATHIAS STECK    Right, actually you gave me a good lead into this topic earlier. So DNV GL to a big extent is also a company doing inspections. And when this whole drone technology came up, we considered this a disruptor to our business. But then when I met you during a presentation at the Hub here in Singapore, it struck me that, in fact, autonomous drones are a disruptor of the disruptive technology, or disruptive drone technology. So, I’d like to understand a little bit more what your view is where drone technology will go in the future, and what that will do actually to the industry. For example,  for companies like ours.

PULKIT JAISWAL    Sure, so, the way I see drones right now I think drones are becoming extremely commoditised as is the case with any hardware product.


The margins are really low, and people keep adding new features over a period of time, but the process of adding those features is extremely slow. And over time, people will simply buy the cheapest drone that’s available. So, it’s basically a commodity, the way I see it. And if I were to draw an analogy I would say that drones are where computers were in the early 70s, when you had IBM running computers for big banks, because those computers were really big and they would often break down. And then the personal computer really revolutionised everything because you did not need many people to run drones for you. And this is the case with pilots right now, pilots operate drones and then you need a safety pilot to make sure that the pilot is doing the right thing. Although drones are disrupting the industrial inspection space, there’s a huge leverage right now for companies like SwarmX to come up and say, “Hey, what if we automate this entire process just like the personal computer operated automated mainframe computers?” and really killed the need to have pilots. And over time, what’s going to happen, is companies like SwarmX will build extremely reliable systems that completely remove the need to have people on the ground. And this is really the disruptor of the disruptor.

MATHIAS STECK    Pulkit, you just mentioned drones are a commodity, and I sometimes wonder, there’s quite a lot of solution providers with drones out there.  What is the real differentiator? Is it the fine stuff or is it rather the dashboard of the data processing in the back?


PULKIT JAISWAL    So, the way we see it hardware is never the defensible part in any company. Hardware is just a means to acquire data for us, and simply a means to project it on a dashboard. What you’re actually consuming is not the hardware, it’s basically infrastructure. The best way to think about is, think about how you make phone calls to your friend. When you make the phone call you do not think about how many times the phone call went through different telecom towers before it reaches your friend.  Instead you just focus on talking to that person, and as long as the data gets conveyed you’re happy. And you pay the telecom tower company, which is your telecom provider, for giving you that service. So, the way we see ourselves is we are simply an endpoint to get that data that’s necessary to help companies make better decisions.  How we differentiate ourselves really, is we really solve the problem of data acquisition. Right now, we still need pilots.We get rid of pilots, we make the data acquisition really smooth. But again, our secret sauce is really how that data is acquired. We have high performance GPUs in our system allow us to do real time data processing. And there are not a lot of companies in this world that can do that in real time, simply because of bandwidth issues, and we’ve solved all of those problems.


And of course, we have a dashboard which we call as the pipeline, and we display exactly the data that you’re looking for, as opposed to displaying the bigger picture. Because at the end of the day you’re only interested in those key pieces of information that help you make better decisions. So, that’s really how we separate ourselves. But again, like I said drones, the base stations, they’re not the long term goal for us.  For us, it’s how quickly we can get the data back, and help make decisions. That’s essentially what we focus on.

MATHIAS STECK    Right, actually, maybe we can talk a bit about this example we tried to pilot together and are about to do together – how to use drone technology, and to perform, or to optimise the performance of assets – and we were talking here in this particular case about a solar PV asset. So, maybe you could just describe this, DNV GL has obviously given this Smarter Operations suite which we offered for quite a lot of clients.  How did you integrate your drone into that particular application?

PULKIT JAISWAL    So, when it comes to solar farms, there are a few key issues that we have identified that solar farm owners would like to look at. One of them is, again, the structural integrity of their asset, which is in this case the solar farm. Is it cracked in a few places? Is there deposit on a certain sector? Especially when the size of the asset gets really big it’s really difficult to manage it.  And often times you have to build scaffolding and have people climb up and inspect them one by one, so it feels like you’re still in the dark ages.


And I would say the same thing for hotspots. Sometimes certain panels get heated up, and this really lowers down the efficiency of solar grids as we know. And those are things that a person with a thermal camera can diagnose and allocate to the nearest operations and maintenance person, Who in turn, again, builds the scaffoldings and goes and fixes that thing. So, although I see that these renewable assets, they’re really changing the world, the way we manage them is not at its optimum level. And drones can really further the case for renewable energy, in this case, for solar farms, by drastically reducing cost and time. And in our case, we see that by having our base stations strategically placed along the infrastructure, we can automate those inspection tasks without having to send anybody at all – and that’s our long term goal. And for the short term, of course, we are trying to validate our technology, it will take many months or many years before we can hit fully autonomous operations. But the Thailand project is really a start for us to start getting that initial data to validate this hypothesis.

MATHIAS STECK    Right, so when I see people looking into digitalisation, or robotics, and probably even also ourselves, we’re looking a lot into this. At the beginning, I think you start thinking, “how do I use this technology now to make my services better?” I kind of came to the conclusion that you maybe got this stick from the wrong end.


The question is rather, “There’s a problem. Can I, with my drone technology, offer the solution, which is the most competitive solution to this?” So, if we open up this question a bit to the overall industry space, are there any applications you can talk about where the drone maybe wasn’t the first solution you were thinking about, because of it being a drone, but where you could really add value because you were quicker with the other things you just mentioned.

PULKIT JAISWAL    Yes, sure, so, the way I see is, drones are basically an alternative. Drones are essentially mobile sensors, and those sensors can be static as well. And one of the industries that we focus on, again, apart from asset monitoring, is security and surveillance. And we’re working with big players like Tyco, Tyco Security, Tyco Innovation. And, what we’ve noticed is that, they’ve already solved maybe, 60% of all problems in the security industry, simply by having these static IP cameras. Low cost IP cameras deployed all around Singapore, Israel or pretty much most of the ASEAN region. And they’ve completely locked in the entire terrain that they want to inspect. So, for them, 60% of the time, they don’t really need a drone. And in this case all they really need is an amazing dashboard, that takes in data from these disparate sensors and displays them, and only shows the stuff that’s interesting. So, this is again where we come in and say, “Hey, you don’t really need a drone. What you really need is our GPU that can take in the drone, capture, take in the static camera feed, run it in real time and only grab the frame of interest, and send it to the command centre.”


So, again we’re not dead on drones. The way we see it is they are just sensors. Those sensors can be static as well. In fact, for our latest appointment with Tyco, we’re actually looking at integrating an ecosystem of IP cameras, with a few drone systems just to start. And over time, when it becomes obvious that mobile sensors are a better option, and they are comparable to the price of static cameras, we will transition fully to drones. So, that’s essentially the way we’re seeing the industry, especially for our case,  take off.

MATHIAS STECK    Right, okay, great. Unfortunately we have already come to an end here with our time, but thank you very much for all the insights Pulkit, and I hope our pilot starts soon.

PULKIT JAISWAL    I’m looking forward to it, Mathias. Thank you.

MATHIAS STECK    Okay, thanks, bye bye.

NARRATOR            Thank you for listening to this DNV GL Talks Energy podcast. To hear more podcast in the series please visit dnvgl.com/talks energy.