🚀 AgroBot is Back – And Now Officially Published with the Indian Patent Office! 🚀
When my team and I first built AgroBot for the Smart India Hackathon 2025, reaching the national shortlisting (top 50 out of 800 teams) was a massive milestone. But we always knew the project was far from over. We took everything we learned, went back to the drawing board, and completely re-engineered our core technology.
Today, I’m incredibly proud to share that our revamped system has been published as a complete specification patent by the Indian Patent Office under Vellore Institute of Technology, Chennai! To make AgroBot more effective, we completely overhauled the machine learning logic. We moved beyond basic image processing and implemented advanced pixel-level segmentation combined with droplet-based processing.
Here is a look at the technical leap behind our newly patented 💫 "CLOSED-LOOP LIQUID DISPENSING SYSTEM WITH IMAGE-BASED SPATIAL TARGETING AND DEPOSITION VERIFICATION" 💫 :
🪴 Pixel-Level Precision: The control unit generates a pixel-level infection segmentation map from captured image data, precisely identifying the geometric boundaries of infected plant regions.
🪴 Spatial Targeting: The system converts this pixel-level infection segmentation map into spatial spray vectors, which include exact nozzle orientation angles and targeting coordinates.
🪴 Micro-Droplet Dispensing: It actuates a servo-controlled nozzle and a diaphragm pump to dispense pesticide micro-droplets directly toward the infected plant regions according to the computed spatial spray vectors.
🪴 Closed-Loop Verification: Through the vision system, it captures post-spray verification images associated with the treated plant regions and determines the actual droplet coverage.
🪴 Autonomous Recalibration: Based on the determined droplet coverage, the system recalibrates its spray parameters, including nozzle angle, pulse duration, and pump activation, to ensure accurate application. This evolution represents a massive leap from our initial SIH prototype to a geometrically precise, self-correcting agricultural solution.
A huge thank you to My Team Eclipse 🌘 , Harshini Subbiah, Harini Subbiah, Lalith Kishore M, LALITHYA KRISHN, and Mirunalini A, and a special note of gratitude to our mentor, Prof. Dr. Manikandan P, for the continued guidance.
#AgroBot #PatentPublished #VIT #MachineLearning #ComputerVision #AgriTech #Innovation #EngineeringJourney #ClosedLoopControl #SmartFarming
26 Julio 2026 à 05h30
Ver en LinkedIn →
