Two years ago, our team built an automated vegetable transplanting machine for our final-year project at the University of Moratuwa. Today I'm a mechanical design engineer, and the biggest lesson from that project came from the part that failed. 🌱⚙️
In Sri Lanka, many vegetable farmers still transplant seedlings by hand: ropes to mark the rows, manure to mark each spot, and long hours bent over the soil. Coming from farming families region, we wanted to change that.
🔧 How it works
Nursery-grown seedlings are loaded onto a hexagonal tray, chosen through weighted concept scoring. The tray rotates and slides on a lead screw to position each seedling over the transplanter, which opens, lowers the seedling and plants it in the soil. The machine then moves itself to the next planting point.
🧭 How it navigates
GPS, an IMU and wheel encoders combined through an Extended Kalman Filter, with four-wheel steering to switch rows. The control system runs on a Raspberry Pi 3B+ and an Arduino Mega, driving 6 motors.
🛠️ What we delivered
✅ A full-scale design (54-cell tray) and a lab-scale prototype (18-cell tray), modelled in SolidWorks
✅ FEA on both chassis to optimise section sizes and bracing
✅ A prototype we fabricated ourselves: lathe, milling, grinding, drilling, welding and laser-cut parts
✅ Seedling-drop accuracy improved to ~83% through iterative testing
👨🔧 My role
I was responsible for the transplanting and steering systems, from research to design, calculations and fabrication:
🔹 Transplanting mechanism: reviewed existing transplanter designs, including a field visit to FMRC, then designed a slotted-arm mechanism that converts motor rotation into the motion needed to open, lower and plant each seedling. I calculated the force needed to penetrate the soil and the motor torque required.
🔹 Steering system: designed the four-wheel 90° steering that lets the machine move sideways between rows, calculated the torque needed to turn each wheel in the soil, and sized the wheels using a practical soil-sinkage test.
🔹 Fabrication: helped build the steering system and transplanter using lathe, drilling, welding and grinding.
💡 The lesson I still use every day
During testing, our transplanter motor burned out. The calculations were correct. The problem was fabrication: the guide bars weren't perfectly parallel, and a welded motor mount was slightly off 90°. Those small errors created loads the motor was never designed to carry.
Now, producing fabrication drawings for heavy-duty equipment, I think about that motor often. A design is only as good as its tolerances and how it's actually built.
Thank you to my teammates Andrew Ferdinan and Rathees Thillainathan, and our supervisor Dr. Dumith Jayathilaka, for an unforgettable journey. 🙏
Engineers: what's a small fabrication error that taught you a big lesson? 👇
#MechanicalEngineering #AgriTech #Mechatronics #SolidWorks #UniversityOfMoratuwa
28 September 2026 à 15h27
View on LinkedIn →
