NeoPangea

Bringing TItanic to life, immersively! ⛴️

Our newest collaboration with E/M Group and RMS Titanic Inc. brings history to life through a rich, educational, and interactive experience.

We’re thrilled to be a trusted collaborator in crafting this journey. We hope the experience will stay with visitors long after the voyage ends! Stay tuned!

Black Friday's MUST-HAVE TOY this year (that nobody asked for!) 🪀

Product visualization tech is so good right now that if your brainstorm doesn’t produce quick iteration, people will ask, “is your Wi-Fi down?"

With AI as a realized process we can: cue instant visuals, cue laughter, cue decisive pivots—and then refine!

For our Black Friday “Must-Have” toy experiment, we gave AI everything it needed to feel alive: reference imagery, a product style sample from the LEGO Group, and a prompt library that helped guide the AI to generate succinct and clear point-of-view images of our headquarters as a buildable limited edition theme set.

Our prompt went something like this:
👨‍💻“Hello AI, our souls are trapped in these offices from 9–5. Could you make it a toy? Blocky, please. Only 4 or 8 colors–blue, yellow, white, red. Place figurines of children on the roof, children love roofs.”

What once required days of design exploration can now be tested with a handful of sentences and a willingness to direct the outcome. That’s the power of AI: working at the intersection of preparation, perspective, and knowledge, where ideas sharpen. With AI, we take more shots—and miss far less!

With guidance from our creative team, AI-generated concepts can render almost pitch-ready, and even its misfires help us test, explore, and uncover challenges before a real opportunity arises. Just ignore the smoke and explosive puttering noises...

So… have an idea to be Black Friday’s hottest toy? Put it through AI and see what happens! But not to spoil the surprise—this year, it’s us!










[Disclaimer: Lego did not endorse, nor did they have anything to do with NeoPangea's playful, experimental R&D]

What happens when touch doesn’t connect?

In the laboratory, we’re building a full scale replica of a touch interface, a controlled testing environment that helps us identify interference, errors, and bottlenecks long before installation begins for an exhibition interactive.

The system, built with Kinect hardware and Python based software, creates a sensor field that detects touch with high precision, essential for top-secret exhibition currently in development. By capturing depth data, the system identifies anomalies between successful touches and errors, ensuring that every interaction is fast, intuitive, and accurate.

When the interactive eventually is fully installed on location, every second will matter. Visitors will expect seamless access to information, so our goal is to design an interface that feels effortless, from trigger to tap. Through precise depth calibration and fingertip tracking, we make sure that when someone points, the system responds exactly as intended. No room for error, not even a fingernail's length!

Our latest prototype uses depth and RGB data from a Kinect sensor to establish a baseline map of the surface. By comparing fingertip positions against that map, the system detects taps when a finger comes within about 2 cm of the surface. If hand tracking fails, a custom depth-based fallback keeps detection consistent. That’s the kind of proactive testing we’re all about (think of it like blasting the system from every possible angle just to see what'll rattle loose!)

While the current version still requires refinement to concrete a final interactive, the proof-of-concept at this stage already demonstrates accurate and reliable touch detection, bringing us one step closer to flawless interaction design!