← Back to selected work

ROBOTICS × BOTANICAL EDUCATION

HerbiQ

A garden guide that connects plants, people and technology.

Status
Team project · WRO India 2026
My focus
Team coordination, AI integration & research
Tools & concepts
Raspberry Pi · ESP32 · ArUco · Node.js
HerbiQ visitor website showing its botanical education home page and Tulsi plant profile
The live HerbiQ visitor website — the digital companion to the robot. Captured October 2026.

01 / THE IDEA

Why this project exists.

A medicinal garden holds a lot of knowledge, but a plant label can only say so much. HerbiQ is Team Delta Malabar’s Ayurvedic heritage guide robot, designed to help visitors discover plants through a touchscreen and spoken interaction.

02 / MY CONTRIBUTION

My part in the process.

As part of Team Delta Malabar, I helped coordinate the project, worked on AI integration and researched the plant information. My work connected the visitor experience with the plant database and the hardware experiments. HerbiQ was a team effort, with software, electronics, mechanical work and presentation coming together.

Team recognition: Silver Badge at WRO India National Level 2026, Hyderabad, for Team Delta Malabar’s HerbiQ project.

03 / THE EXPERIENCE

What it brings together.

Identify the plant

The demonstrated workflow uses ArUco markers to identify a plant and retrieve its local record.

Ask and explore

A touchscreen, local plant information and multilingual voice interaction make the content approachable.

Connect hardware and software

Raspberry Pi handles the interface and plant data; ESP32-based hardware supports sensing and motor control.

Continue on the web

The visitor website offers plant browsing, QR discovery and AR-ready plant viewing.

04 / BUILD NOTES

The difficult parts.

The difficult part was getting several small systems to work together: voice input, the plant database, the display and the robot hardware. During development, storage instability interrupted Raspberry Pi work, and an ESP32-CAM hardware repair was needed before image capture could continue.

I worked through individual components before joining the full system. English and Malayalam voice input/output were tested, and an offline chat experiment connected the Raspberry Pi interface to a phone-hosted model over local Wi-Fi.

05 / WHAT I LEARNED

Lessons for the next build.

  • A working identification method matters more than an ambitious feature list. ArUco markers gave the garden interaction a clear starting point.
  • Hardware reliability and data backups are part of the product, not tasks to leave until the end.
  • Short, readable plant records and clear voice responses make the information easier to explore.

06 / STILL EXPLORING

Where it goes next.

The next focus is making the visitor experience more reliable and expanding the educational content. Direct visual recognition and plant-health analysis remain exploration areas; they are separate from the demonstrated marker-based workflow.

KEEP EXPLORING

Kerala Routes ↗
All projects ↗