GeoBox: Augmented Reality for Geoscience

One of the first challenges geology students encounter in the classroom is learning how to read a topographic map. With contour lines, a map can transform a flat sheet of paper into a 3D representation of landscapes with rivers, valleys, and mountains taking shape through imagination. Yet, for many students, truly visualising Earth’s processes across space and time remains difficult.

This is where Augmented Reality (AR) has the potential to change geoscience education.

What is Augmented Reality?

Augmented Reality (AR) superimposes digital elements like images, colours, and even simulations on top of the real world. The most popular examples of this technology could be the trending game from 2017 “Pokemon Go” and the annoying Snapchat filters. Unlike Virtual Reality, which immerses you in a fully artificial environment, AR enhances your existing surroundings. With just a camera, and projector, digital and physical landscapes blend into one seamless experience.

The Promise of the AR Sandbox

For context, an AR Sandbox is essentially a box of sand combined with a 3D sensor (like a Microsoft Kinect) and a projector. The sensor scans the sand’s surface in real-time, sends the elevation data to a computer which processes this information and creates a visual output and the projector overlays the visual output containing contour lines and colours that shift as you reshape the landscape.

Many in the geosciences community are already familiar with the AR Sandbox, first developed at the University of California, Davis by Oliver Kreylos. This interactive tool projects topography onto a real sandbox, allowing users to shape terrain by hand and instantly see contour lines and water flow simulations come to life.

It’s an incredible concept: students can sculpt mountains with their hands, carve valleys, and watch how landscapes evolve in real time. For undergraduates and even high school learners, it turns abstract map reading into an intuitive, hands-on experience.

But there’s a catch. Typical AR Sandbox setups remain:

• Expensive, especially when purchased from vendors. Commercial setups range from $7,990 to over $25,000, with software licenses alone running into thousands.

• Complex, requiring technical expertise in programming languages like C/C++.

• Rigid, the software architecture in most setups often lack modularity, making customization difficult.

• Not geoscience-focused: Most existing versions are general-purpose, with little emphasis on geoscience applications.

These barriers restrict its widespread adoption in classrooms especially in resource-constrained institutions.

The Mission: A Smarter, More Accessible Sandbox

Recognising both the strengths and limitations of the AR Sandbox, I wanted to “rebuild” the AR Sandbox for Geosciences into the ‘GeoBox’.

GeoBox is designed to make AR-based geoscience education affordable, modular, and simple. We’re building it so that:

• Hardware uses equipment most institutes already own.

• Setup and calibration are streamlined through a desktop app which means no coding required.

• Thanks to the software’s architecture, if you know your way around in Python, modules can be added or customised depending on the teaching focus.

Why Python?

We chose Python as the backbone of our system. It’s readable, modular, and widely used in both education and data science. This means:

• Faster development with less code.

• Easy integration with scientific libraries.

• A modular design where researchers can plug in their own tools and simulations or create their own modules.

In short, Python makes our sandbox future-proof and collaborative.

The Journey So Far

The first prototype of GeoBox is already operational at the Department of Geology, St. Xavier's College, Mumbai. During its initial testing phase, the sandbox ran on existing platforms like the SARndbox software, followed by Open_AR_Sandbox (a Python-based project by Florian Wellmann) which was later ported to Unity. For demonstrations, we also experimented with Magic-Sand (developed by Thomas Wolf).

These experiences offered us valuable insights into the strengths and shortcomings of different software approaches. Today, GeoBox has its own skeletal version in place, steadily evolving into a self-contained ecosystem tailored for teaching geology.

By February, I plan to roll out new geoscience-focused modules including hydrology simulations.

Looking Ahead

The vision is clear: take advantage of AR in geoscience education. By making the technology modular, affordable, and intuitive, GeoBox can open up opportunities for schools, colleges, and field labs that would otherwise find AR tools out of reach.

The journey is still ongoing, but the potential is enormous.

What features would you like to see in the GeoBox?