RoboClub Academy · Student Robotics Team
Arnav Chaudhry · Vedaang Aggarwal · Yaj Mehra — mentored by Kritik Bhatia

THE MECH
MAHARAJAS

Our Journey to Build Heritage Guardian
Three students.
One problem.
One robot.
One mission: protect our heritage.
Explore Our Journey → Meet Heritage Guardian
SCROLL
Our Story

Not just a robot — a question we couldn't let go of.

We started our robotics journey through RoboClub Academy, guided by our mentor Kritik Bhatia. Instead of picking a random robot to build, the three of us — Arnav, Vedaang, and Yaj — wanted to find a real problem worth solving.

That search led us to India's heritage — and to a question that kept surfacing every time we read about conservation work: how do you watch over something too large, too old, and too fragile for any team to observe all at once? The more we researched, the more it looked like a problem robotics could actually help with.

We saw a problem
We researched it
We questioned how tech could help
We imagined a robot
We built
Things went wrong
We learned
We improved
Heritage Guardian evolved
The Problem

Watching over a monument is harder than it looks.

India has an enormous number of historical and cultural heritage sites — many of them vast, intricate structures with walls, corridors, pillars, carvings, ceilings, and courtyards that all need attention. During our research and site visits, four challenges kept coming up.

01 — STRUCTURAL

Cracks & Structural Damage

Cracks can develop in walls and other parts of a monument. Small cracks are easy to miss during routine observation, especially in less accessible areas — and damage that isn't documented early becomes harder to monitor and manage over time.

This is one of the central purposes of Heritage Guardian: helping detect and document visible signs of damage so human experts can investigate.

02 — SCALE

The Difficulty of Continuous Monitoring

Large monuments contain many rooms, corridors, walls, pillars, and surfaces. Human teams simply cannot stand everywhere at once. A robotic platform that moves through accessible areas could collect information more regularly — as an additional monitoring tool, not a replacement for people.

03 — ENVIRONMENT

Environmental Degradation

Heritage monuments are exposed to their surroundings for years. Conditions can contribute to deterioration over time, including:

  • Temperature changes
  • Humidity & moisture
  • Pollution & dust
04 — HUMAN ACTIVITY

Damage Caused by Human Activity

Heritage sites can also suffer damage from human activity, including:

  • Graffiti and scratches
  • Dents or physical contact
  • Improper handling of surfaces
Our Solution

Heritage Guardian: technology-assisted monitoring, not replacement.

Heritage Guardian is a mobile heritage-monitoring robot. It is not designed to replace archaeologists, conservation experts, or restoration teams — it's designed to assist them, by moving through accessible areas, collecting data, and flagging what deserves human attention.

A / MOVE

Navigate the site

The robot can move around a heritage environment and reach accessible areas.

B / SENSE

Read the environment

Sensors collect information about surroundings and environmental conditions.

C / OBSERVE

Inspect surfaces

A camera/vision system inspects surfaces and collects visual information.

D / DETECT

Flag potential damage

AI/computer vision supports a developing crack and damage detection system.

E / REPORT

Surface the data

Collected information can be displayed or sent to a monitoring system for human review.

F / INTERACT

Engage visitors

The robot can talk to visitors and explain why heritage preservation matters.

What Heritage Guardian is not: it cannot restore monuments, repair structural cracks, or guarantee that every flagged line is a real crack. It cannot replace conservation experts or safely enter every part of every site, and it is not currently deployed at any monument. It is a student-built prototype — a monitoring and early-warning assistant, with every detection intended for human verification.
How It Works

From heritage site to human decision.

Every scan follows the same path — and it always ends with a person, not the robot, making the call.

Step 01

Heritage Site

The starting point — walls, corridors, pillars, and surfaces that need watching over.

Step 02

Heritage Guardian moves through accessible areas

The robot navigates the parts of the site it's able to safely reach.

Step 03

Sensors + camera collect data

Environmental readings and surface imagery are gathered as it moves.

Step 04

Raspberry Pi / Arduino process information

Onboard computing handles sensor input and prepares data for analysis.

Step 05

AI / computer vision analyzes visual data

Captured images are examined for visual patterns that may indicate a crack.

Step 06

Potential cracks / changes / conditions identified

The system flags what it observes — described as potential, never confirmed.

Step 07

Data displayed / recorded

Findings appear on the robot's display and are logged for review.

Step 08 — Human in the loop

Human expert reviews the result

Every flagged result goes to a person before anything is acted on.

Step 09 — Human in the loop

Conservation team takes appropriate action

Final decisions and any physical work remain entirely with human experts.

Crack Detection & Baseline Comparison

Camera → image → AI → potential crack → human review.

A baseline is an earlier reference image of a surface. When Heritage Guardian scans the same area again, the new observation is compared against that baseline to help identify changes worth a closer look — not to declare a diagnosis on its own.

Potential crack flagged — for review
BASELINE
CURRENT SCAN
Drag to compare — [ADD REAL CRACK DETECTION DEMO / SCAN FOOTAGE HERE]

The system does not claim certainty. A flagged line means "this looks different from the baseline and may be worth a human look" — nothing more.

Technology

What's actually inside the prototype.

Tap a component to see how it fits into Heritage Guardian. This is our real hardware — nothing here is invented or aspirational.

+Computing

Raspberry Pi 4

The main computing platform — handles processing, camera and computer-vision tasks, AI-related processing, data handling, display output, and communication between systems.

+Microcontroller

Arduino Uno R3

Handles low-level hardware: sensor input, servo control, and LEDs. Works alongside the Raspberry Pi, which takes on the more computationally intensive tasks.

+Interface

7-inch HDMI Display

The robot's main interface — shows status, sensor readings, heritage information, alerts, detection results, and visitor interaction.

+Sensor

HC-SR04 Ultrasonic Sensor

Detects distance to nearby objects — helping the robot sense whether a person or object is close, supporting navigation and interaction.

+Actuator

SG90 Servo Motor

Drives small mechanical movement, such as a robot arm or mechanism. We've experimented with triggering the servo in response to distance detection.

+Display

I2C 16×2 LCD

Used in our electronics experimentation to display simple information like sensor readings or system status.

On the AI side: computer vision is used for a developed / in-development crack-and-damage detection concept — every result is described as a potential detection for human verification, never a confirmed structural diagnosis.
Visitor Interaction

Heritage Guardian doesn't just watch — it talks.

When a visitor approaches, the proximity sensor can detect them, and the robot responds through its display and audio — sharing why the site matters and what Heritage Guardian is doing to help protect it. Interaction is designed to happen in Hinglish where appropriate.

"Namaste! Main Heritage Guardian hoon. Main heritage monuments ko monitor aur preserve karne mein help karta hoon."
— one example of the interaction style, not a fixed script
Our Journey

Ten stages. A lot of failed wiring. One robot.

Tap any stage to open it. This wasn't a straight line — and we're not pretending it was.

01

The Beginning

›

We started our robotics journey through RoboClub Academy with our mentor Kritik Bhatia. The three of us — Arnav, Vedaang, and Yaj — began exploring robotics, electronics, programming, and real-world problem solving together.

02

Finding a Real Problem

›

Instead of starting with a random robot idea, we wanted to identify a real problem. We researched India's heritage and the challenges involved in protecting historical monuments — which pointed us toward the problem of monitoring large heritage structures.

03

Heritage Research

›

We researched heritage monuments and preservation challenges, and visited heritage sites to observe issues firsthand — cracks, damage, environmental degradation, human activity, and just how hard it is to monitor a large area continuously. That's what turned an abstract idea into a real problem.

04

Brainstorming

›

We began asking: "What if a robot could help monitor these monuments?" We explored ideas involving sensors, cameras, robotics, AI, monitoring, displays, and data collection. Eventually, the concept of Heritage Guardian took shape.

05

Designing the Robot

›

We started thinking through how the robot should actually work — mobility, sensor and camera placement, electronics, the display, computing, visitor interaction, and heritage-safe operation. The design evolved over time rather than arriving fully formed.

06

Building

›

Hands-on prototype building: wiring, electronics, Arduino, Raspberry Pi, sensors, motors and servos, the display, programming, and mechanical construction.

[ADD ROBOT BUILD PHOTOS]
07

Testing

›

We tested individual systems — distance detection, servo movement, LEDs, display output, sensor readings, electronics, and robot movement. Testing showed us clearly what worked and what didn't.

08

Failure and Iteration

›

Robotics rarely goes right the first time: build → test → fail → understand → modify → test again. We're not pretending our journey was perfectly smooth.

What Went Wrong?

This space is reserved for our actual failures — what broke, what we learned, and how we fixed it.

[ADD REAL FAILURES, PHOTOS & SOLUTIONS HERE]
09

AI and Crack Detection

›

We explored how AI and computer vision could be incorporated into Heritage Guardian — analyzing images of surfaces for potential cracks or visible damage, and comparing new observations against a baseline to identify changes that may need attention.

10

The Final Concept

›

Heritage Guardian: a robotic platform designed to assist with heritage monitoring, environmental observation, visual inspection, potential crack detection, data collection, and visitor interaction.

Our Team

Three students. One mentor. One mission.

Individual roles and contributions will be added here as the team defines them.

[ADD TEAM PHOTO — ARNAV CHAUDHRY]

Arnav Chaudhry

[ROLE TO BE ADDED]

[Individual contribution to be added.]

[ADD TEAM PHOTO — VEDAANG AGGARWAL]

Vedaang Aggarwal

[ROLE TO BE ADDED]

[Individual contribution to be added.]

[ADD TEAM PHOTO — YAJ MEHRA]

Yaj Mehra

[ROLE TO BE ADDED]

[Individual contribution to be added.]

[ADD MENTOR PHOTO — KRITIK BHATIA]

Kritik Bhatia

MENTOR · ROBOCLUB ACADEMY

Guided the team throughout the research, design, and build process.

Arnav + Vedaang + Yaj + Kritik Bhatia
↓
Research
↓
Ideas
↓
Building
↓
Testing
↓
Failure
↓
Improvement
↓
Heritage Guardian
Build Log

Photographs and notes from prototyping, wiring, and testing.

◫[ADD WIRING PHOTO]
◫[ADD ARDUINO SETUP PHOTO]
◫[ADD RASPBERRY PI SETUP PHOTO]
◫[ADD SENSOR TESTING PHOTO]
◫[ADD SERVO / MOTOR TEST PHOTO]
◫[ADD DISPLAY OUTPUT PHOTO]
Where We Want to Go Next

Heritage Guardian today is a prototype. Here's where we hope to take it.

Everything below is a future possibility we're exploring — not a current capability.

→

Improved AI crack detection and computer vision accuracy

→

More accurate environmental monitoring

→

Cloud-based data storage

→

Historical scan databases

→

Automated comparison of observations over time

→

Improved navigation and better mapping of heritage sites

→

More advanced visitor interaction

→

A dashboard built for conservation teams

→

Expanding the system to more types of heritage sites