Titanic Digital Twin: Exploring the Reality Capture Process

Lidar News Revisits the Titanic Digital Twin

Today, Lidar News is diving deep on a subject that has fascinated the world for over a century — the Titanic. Last month, we covered the stunning Titanic digital twin created by Magellan, which captured the wreck in unprecedented detail. This project has already shifted the historical understanding of Titanic’s final moments, revealing features and structural changes never documented before. These revelations were the subject of a recent National Geographic documentary.

But many in the reality capture community were left wondering: how exactly was this Titanic digital twin created? Some media outlets referred to the process as “scanning,” and a National Geographic article stated it involved “millions of laser scans.” Beyond that, there were few details regarding the Titanic reality capture process and the role lidar may have played.

So, Lidar News reached out to Doug Scott, Project Manager at Magellan, who led the Titanic mapping effort, to learn what really went into this monumental project — from hardware and navigation to data processing and underwater photogrammetry.

A 16-Terabyte Undertaking at 3,800 Meters

Capturing the Titanic — which rests nearly 3,800 meters below the surface — was no small feat. Magellan deployed two custom-built ROVs, Romeo and Juliet, to systematically cover the wreck. These ROVs, the size of small cars, weigh roughly four tons each and are tethered to the ship with cables supplying both power and fiber-optic data transfer. Equipped with high-resolution camera systems, the vehicles maintained a delicate 2–3 meter distance from the wreck — close enough for detail, but never making contact, in adherence to strict protocol.

Mounted on integrated skids developed by Voyis (a Canadian-based company specializing in deep-sea observation), each ROV carried a suite of sensors: HD cameras, lidar, strobes, inertial navigation systems (INS), Doppler velocity logs (DVL), and ultra-short baseline (USBL) for acoustic positioning. Lidar was used alongside the camera system throughout data collection, but to date, lidar has not been used in the Magellan reconstruction. Sonar, meanwhile, supported navigation rather than 3D modeling. This was Magellan’s first project incorporating lidar, and Scott expects it to play a growing role in future deep-sea reality capture efforts.

Precision Without GPS: Navigating the Depths

One of the biggest challenges in producing the Titanic digital twin was achieving the right image overlap — Magellan aims for at least 75% — without the benefit of GPS. Underwater positioning relied on USBL combined with INS and DVL. Even with these tools, real-time feedback is difficult in these conditions. Positioning systems subsea are not nearly as accurate as GNSS in terrestrial systems, leading to drift. SLAM was not used on this project as it struggles to handle the enormity of the Titanic site. Returning to fixed points to recalibrate is a major efficiency challenge.

To deal with this, dedicated photogrammetry experts were onboard, providing feedback and guiding ROV pilots to revisit key areas as needed. These efforts were crucial in building a Titanic digital twin that would stand up to scientific and historical scrutiny. More broadly, this phase of Titanic reality capture required an extraordinary level of coordination between navigation systems, human oversight, and imaging tools. ROV operations continued nearly non-stop for three weeks — an extension beyond the original two-week plan thanks to excellent weather and equipment performance.

Processing: The Marathon Behind the Scenes

The dataset ultimately included 715,000 high-resolution images, with the final model totaling 16 terabytes. Processing began onboard, with early-stage quality control to ensure no major areas were missed — critical, given the site’s remote location. Scott estimated the site is at least two weeks’ travel from the vessel’s normal operating regions, and operating costs hover around $100,000/day. It took 3–4 months to generate the first draft of the model, followed by more than a year of reprocessing with improved workflows based on lessons learned.

Magellan’s biggest takeaway from years of subsea photogrammetry expertise? Attention to detail is everything. Coverage, visibility (i.e., low turbidity), and lighting are essential for successful photogrammetry. Achieving that underwater requires experienced pilots who avoid disturbing seafloor sediments and follow pre-defined grid paths with precision — a discipline that ultimately enabled the creation of a highly accurate Titanic digital twin.

Bringing the Titanic to the Public

While the full-resolution model remains too large for general release, a simplified version powers Magellan’s vROV experience, available for $24.99, letting the public explore the wreck in an interactive 3D environment. The company hopes to eventually offer access to the full-resolution dataset, but at this time it is prohibitively large. They also plan to expand the vROV platform to include other underwater experiences, like geothermal vents and other historic deep-sea sites.

For Magellan, this was a passion project. The team happened to be transiting near the Titanic site between projects, and the company’s founder seized the rare opportunity to document the site in an unprecedented way. In doing so, they’ve created not only one of the most remarkable digital twins in existence, but also a new way for the public to engage with deep-sea environments.

We would like to thank Doug Scott and Magellan for sharing their story with us, and we look forward to reporting on other exciting projects in the future!

All photos were provided courtesy of Magellan. Please engage with Magellan using the links below:
– Magellan
– Patreon
– YouTube
– vROVPilot