Pranay Deep
The CS Guy vs. The Laws of Physics (A Story about the project)
So here's the deal: I’m a Computer Science student who spends way too much time obsessing over robotics hardware. I bought an educational "Acebott" robotic arm kit online just to practice writing IoT code on an ESP32. I figured I'd write a few lines of C++, make it pick up a block, and call it a day.
Physics had other plans.
When I actually tried to make the arm lift a decent industrial payload, the tiny MG90S micro-servos in the lower joint completely gave up. They just didn't have the torque. Instead of doing some quick jugaad (a makeshift hack) with tape or rubber bands, I decided to engineer a proper, industrial-grade solution.
The plan? Rip out the weak motors and swap them for heavy-duty, metal-geared MG996R servos.
The problem? The new servos were way too big for the factory plastic chassis. Also, I didn't know how to use traditional CAD software like AutoCAD. But I do know how to code. So, I grabbed my calipers, took some physical measurements, and wrote OpenSCAD scripts (which is literally just 3D modeling using code) to generate brand-new, heavy-duty custom mounting plates and a 3D-printed rotating base.
Once the hardware was rock-solid, I went all-in on the software. I built a custom kinematic physics engine so the heavy arm smoothly accelerates and brakes without tipping over. Then, I hosted a complete, mobile-responsive HTML/JS Web Dashboard straight onto the ESP32's internal memory. Add in a TCS34725 sensor, a Euclidean-distance color matching algorithm, and an 11-step collision-free routing system, and my little educational kit accidentally evolved into a full-blown autonomous industrial sorting cell.
Proof that sometimes, a hardware failure is just an excuse to write better software!