Retelling the story with geospatial data

Few people have pushed the boundaries of spatial data, reality capture and 3D storytelling quite like Pete Kelsey. Throughout his career, he has helped archaeologists, engineers, documentary makers and public agencies bring to life invisible worlds, from shipwrecks to historic sites like Alcatraz. Kelsey translates these complex environments into compelling narratives because he believes great stories change minds, and spatial data is what makes those stories real. In this interview held at Intergeo, he effortlessly jumps from lunar coordinate systems and emerging trends like Gaussian splatting to cautionary tales of reckless drone flights, and reminds us that the real magic happens when the industry’s ‘cool tech’ is used to tell a story worth retelling.

You’ve described yourself as a “spatial storyteller”. What does that mean to you, and how has that perspective shaped your professional approach?

I discovered 20 years ago that, for me, reality data is the most valuable data type of all, because there’s no better medium for providing context. It transcends language. If we didn’t speak the same language, I could show you a model, and you’d say “I get it. That’s my home. That’s my town”. That led to my natural evolution into storytelling. It changed many conversations from “Here is the design, here is the data,” to “Come with us, put on the goggles, we’re going to go in there together”. It’s literally when my professional life changed from black and white to colour.

You’ve worked across archaeology, engineering and entertainment. These fields don’t often overlap, so what connects these worlds in your work?

What all of these domains have in common is a story – and stories are what change minds. If I’m an engineer and you’re the client, I’m really sharing a narrative that I hope you’ll connect with emotionally. You should walk away thinking “I like this, I understand it, I believe in it”. That’s the common denominator - or connective tissue – at the highest level.

Moreover, people don’t want raw data; they want solutions. At events like Intergeo, the reality is that most people don’t care about the hardware or software itself, they care about what it allows them to do. And the senior decision-makers will ask a simpler question: “Is this good for us or not?”. I learned early in my career that a product is only one piece of a larger answer. On the complex, high-profile projects I’ve worked on, it has never been one vendor or one tool; it’s a combination. The engineering challenge is understanding the problem and assembling the right mix. I believe audiences grow when companies focus on being part of a solution, not just selling a product. That’s why I prefer a more strategic approach. Instead of a traditional sales pitch, come with me to a site – like Alcatraz – and I’ll show you the real-world impact.

In the documentary series Drain the Oceans, you helped visualize shipwrecks and underwater landscapes that had never been seen before. What was the biggest challenges in turning invisible seafloors into compelling visual stories?

Well, the marine environment is incredibly challenging. All sensors behave completely differently. Photogrammetry above the water is challenging enough, but photogrammetry underwater is a completely different animal. Turbidity in the water creates shadows. Underwater lighting can help, but if the shadows aren’t consistent, the model is going to look bad.

When working on the wreck of the Arizona at Pearl Harbor about ten years ago, we were using an underwater Lidar system designed for oil and gas infrastructure at great depth. The unit was enormous – about 300 pounds (approx. 135kg, Ed.) – so as a diver, trying to maintain buoyancy while moving it with lift bags, all without damaging the Arizona, was no small task. On top of that, we needed an extremely powerful laser for underwater operations, which made it genuinely dangerous; direct exposure could blind you.

Despite the time limitations, depth limitations and challenging weather conditions, the rewards can be spectacular – like the recent photogrammetry mesh of the Titanic, or images of Shackleton’s Endurance, which some friends of mine worked on. I’m still upset I didn’t get asked to go along on that one! The fidelity is unbelievable and, as I said earlier, it allows us to bring these wrecks to life for people. They can even be afraid of water – as long as they’re not afraid of VR headsets!

(Text continues below the image)

In Wyoming’s Jackson Hole, you led a project that used drones and beta mapping software to create a detailed 3D model of a steep backcountry snowboarding course for Red Bull’s Natural Selection Tour. What did that experience teach you about the evolving limits of drone mapping in extreme environments?

When I did that, it was before terrain-following flight planning software was available, so I flew that with my thumbs. It was irresponsible but it worked, because it’s a one-to-one slope. I’m trying to keep it above the trees and the rocks, while a friend right next to me – with better eyesight – was calling things out. To be honest, it was reckless, but we got very lucky; we were successful in the model. It was called the Natural Selection Tour because the athletes could select their route down the mountain based on our data. That was at altitude just before it started snowing, so the drones behaved differently. I remember the mesh. If you haven’t seen the video, the mesh is quite good!

You were among the first to test Esri’s Terrain Follow feature in a live commercial project. How did that moment of real-time experimentation shape your view of innovation in geospatial technology?

I realized that flying in extreme terrain no longer had to feel risky. The real question becomes: how reliable is it? As long as you trust the underlying terrain data, you can simply monitor the flight. Everyone assumes you’re a brilliant pilot, while in reality, it’s the data doing the work!

On one project, we first flew a small drone to capture photogrammetry and generate a DSM (digital surface model, Ed.). We then used that DSM in the flight plan for a larger drone carrying Lidar, and it worked perfectly. In that moment, I understood how this capability would change my service offerings. Once you have done it successfully, you know how to repeat it, and it becomes surprisingly straightforward. And even a short head start like that can give you a real advantage over other service providers.

Many of your projects bridge scientific accuracy and emotional storytelling. How do you find the right balance between data precision and narrative impact?

At this stage in my career, I look for passion projects. I chase a good story first – that’s the number-one criterion. Based on my past experience, nine out of ten clients do not understand the long-term value of geospatial data. So when I find a great story, I will say “Tell me about your existing spatial assets,” and nine times out of ten the client will reply “What? Spatial what?”. That happened on Alcatraz. So we start from zero. Then I will squeeze all the accuracy and precision out of the time and budget allowed, because I know this data will have a lifespan that will far outlive me.

Over the years, I have grown pretty adept at educating clients on the value of spatial data. I remind them that environments are dynamic, affected by geology, time, the elements. I say: “Do you agree your site will change over time? Yes? Then let us establish a baseline, a single source of truth that future surveys can be measured against for change detection purposes.” “You can do that?” they reply. “Yes, we can, and we have to,” I say.

In my opinion, no one should be capturing data that isn’t spatial, because you cut the life cycle of that data by 80%. In time, it becomes a pretty picture that you can do some interesting things with, but you are selling yourself short. So that’s my advice for all your readers!

You’ve introduced geospatial technology to mainstream audiences through television and documentary projects. How important is that outreach in showing the wider world what our field is capable of?

To me, it is everything! It’s one thing to create a beautiful model with a short life cycle, but how about a model that can live forever? I was talking with some of the Alcatraz team this week – our fingerprints are on that data permanently. That’s why, as I mentioned, if it’s not geospatial, it doesn’t matter to me. That is one part of the message. The other part is that this industry is exploding, yet there is a real shortage of talent. It’s one of our biggest challenges. I am ageing out of this – I have been doing it for 30 years – so when I can, I go into schools and say “Come with us. Let me show you how this works, because you’re going to think it is pretty cool”.

There is a bit of evangelism in that. I never had children of my own, so what is my legacy? Well, we need more people in this field. And the reality capture industry is fun – you get to fly drones and use amazing technology, lots of gadgets. The ‘wow factor’ is addictive. And when a young person is impressed by something you’ve done, you can say “I can show you how to do it”.

We should ask ourselves what the industry can do to attract the younger generation, perhaps even at the school level: internships, free software, outreach. Many large vendors already offer free tools if you have an email address ending in .edu, and I love that. Hardware is trickier because it is expensive, but it would still be a smart investment for everyone in this space. Create your future customers. Go into a school, take drones and scanners, and evangelize.

The geospatial sector – and especially the field of reality capture – is evolving rapidly. Where do you see the most exciting developments happening, and how might they redefine what’s possible in the next few years?

Well, in the last 24 hours here at Intergeo, I’ve discovered that I can no longer dismiss Gaussian splats. As a geospatial guy, I’ve always thought they are beautiful. The datasets are spectacular. But there has been one problem: they are not spatial, and they are not measurable. But who better than Esri – who helped with the photogrammetry for Alcatraz and did a beautiful job – to recognize that limitation and say, “We’re going to fix that”.

The ‘aha’ moment for me came when I saw my friend Jeremiah from Esri at their booth, and he said, “Hey, I want to show you something.” Esri has clearly doubled down on reality capture, and he showed me a splat of a city and it was stunning. He knows me well, so he said, “Watch this”, picked a point, and showed northern/easting, elevation, the description, and then started measuring distances. I just looked at it and thought “Alright, I cannot ignore this anymore”. So going forward, I will be paying very close attention to Gaussian splatting as it evolves. It’s time for this old dog to learn some new tricks!

The acid tests, of course, are land surveyors and civil engineers – and that is where I come from. If they start saying it is good, then it’s good. And based on what I saw, I think that is coming in the very near term.

Besides that, I have probably seen 30 or 40 handheld Lidar systems here (at Intergeo, Ed.), and many now have RTK receivers on board. That has been important in my own evolution; it adds credibility and creates a unique service offering for clients.

What do you see as your main personal drive – the motivation that keeps you exploring, experimenting and telling stories through data?

Well, as I have said, I love a good story. That is number one. But I also love a challenge. If a project does not have an immediately obvious solution, I lean into that. I guess it’s in my DNA. But I don’t do any of these things alone. It’s about putting teams together – with people smarter than I am – and solving a problem that others said would never work. It takes a village to solve some of these problems. Fortunately, after 30 years, I’m surrounded by the right people.

And I know it can have a positive impact, even greater than what we did at Alcatraz. No one had tried that combination before: SLAM Lidar, terrestrial laser scanning, drone-based Lidar, photogrammetry, multispectral imaging. It felt like witchcraft, but we figured it out. That feeling, as golfers say, is like sinking a long putt. It’s what keeps you coming back.

After years of capturing the world’s most challenging environments, what continues to fuel your curiosity? And where do you see the next frontier for spatial storytelling?

I have done maritime, terrestrial and underground work, but I think the next great environment will be extraterrestrial. This speaks for itself in terms of the investment required. Elon Musk, Blue Origin, Bezos – these are people who will ask the right question: do we have spatial data of the Moon or Mars? And if the answer is no, they would be out of their minds to even try to go there, let alone build anything or choose the right location.

But some friends of mine are already working on this. They are actively modelling both Moon and Mars missions. So, a coordinate system for the Moon? Yes. Wow. Figuring all of that out also touches on classic AEC (architecture, engineering and construction, Ed.) workflows, which is my background.

We have had fascinating conversations about the fact that we cannot possibly transport all the building materials to another world – it is simply too expensive. So we will need to do what we have always done: look for caves. Then you only have to build one wall and one door. Send up Lidar drones and robots to survey those caves, model them all and then make an informed decision about which one is best for the mission.

The first major project I myself ever did was on Easter Island – which is almost cheating, because every kid knows the famous heads. I scanned them, intentionally, because if you start with subject matter like that, people will pay attention. And because we created a single source of truth, we can go back. Those scans were done 18 years ago. Now, we’re talking about scanning them again, measuring erosion rates for change detection. This brings us back to what I said: lead with a great story. But beneath all of it are real problems we can solve: elevation which is a classic AEC concept, sea-level rise at Alcatraz, or where could we build a habitat on Mars. At the centre of it all is spatial data.

To finish this interview, I’d like to ask you about Alcatraz. What are the main lessons you learned from that project?

I learned that not having a box to think outside of is not as scary as I once thought. If you surround yourself with very capable, very smart people, the problems start to fall away. Personally, it’s affirming.

Everybody has heard of Alcatraz. That choice was very deliberate. I had done a lot of work with the National Park Service in the US because they have such great content and such great stories. But Alcatraz is small – only 23 acres; it’s not the Grand Canyon. It is manageable, yet it has global recognition. I have joked about this for years: Easter Island, Alcatraz, dinosaurs, space, UFOs – those are the subjects everyone pays attention to.

So now I am thinking about the next great story. These ideas are purely conceptual for now, but I am pitching them to people I know. What sits on that same level as Easter Island and Alcatraz? Vampires. The legend of Dracula. Bram Stoker based his novel on actual physical places. Yes, there are castles in Romania, but there is also a site in the north of the UK where he first came ashore, saw an old ruined abbey and thought “I have an idea for a book.” Let’s go scan that. Dracula? Who would say no?

Another one I’m excited about is King Arthur. He was a real historical figure, and the UK is full of sites associated with him. Let’s use geospatial technology to differentiate fact from legend, and at the same time expose people to what this technology can do. It’s a gateway drug to what you and I do for a living. Friends now ask me “What’s next? Where are we going?”, and that affirms the power of great stories. They open doors!

About Pete Kelsey

Pete Kelsey has spent his career championing digital preservation and pushing the boundaries of the geospatial and AEC industries. Above all, he is a geospatial storyteller and one of the field’s foremost evangelists, known for his sharp vision of what the ultimate reality capture toolkit should be. With a background in civil engineering, land surveying and remote sensing technology, Kelsey has led strategic projects around the world documenting historic, cultural and archaeological sites and artifacts. As the owner of VCTO Labs, he helps filmmakers and other creatives bring stories to life through cutting-edge 3D and geospatial technologies. Kelsey’s work has provided photorealistic, survey-grade data that lends accuracy and depth to documentary series such as Drain the Oceans and Buried Secrets of World War II.