"Where AI Meets the Aisle"
← Back to PortfolioGlideX is a compact, intelligent robot built to transform the way warehouses operate. It blends automation and design to create an intuitive solution that saves time and reduces effort. With a strong focus on efficiency and ease of use, GlideX takes on the hard work—so you can focus on what matters. It’s robotics made practical.
Precise, omnidirectional movement — perfect for tight spaces and smooth maneuvers in real-time.
Real-time mapping + localization using YDLIDAR T-mini Plus for autonomous exploration and pathfinding.
YOLOv8 + OpenCV to identify, verify, and classify items in real time for seamless automation.
Flexible arm replaces conveyors — picks and places packages with AI-guided precision.
Flask-based control UI lets you monitor, dispatch, and control GlideX from any browser, in real time.
Mounted interface for local manual control, system status, and dispatch from warehouse floor.
Prototype design & component selection
Hardware integration + motion testing
SLAM & navigation logic complete
Object detection & AI integration
UI dashboard + system testing
Last updated: March 31, 2025
Over the past few months, I’ve been working on building something I’ve always wanted to bring to life — an autonomous robot that could handle warehouse operations like product pickup, sorting, and dispatch, all by itself. That’s how Glide X was born.
Glide X is a 4-wheeled robot designed to navigate like a pro, recognize objects using AI, and physically interact with its environment using an actuator and a robotic arm. It’s meant to automate the entire process of picking up a product, verifying it, loading it, and dropping it off at the dispatch zone — all based on commands sent from a web interface.
The robot moves on Mecanum wheels, which allow it to move in any direction — forward, backward, sideways, and diagonally — making it perfect for tight warehouse spaces. The control system is handled by a Raspberry Pi 5 (8GB), powered safely using a 4S2P 14.8V Li-ion battery pack and regulated with DC-DC step-down converters and a UBEC for 5V power.
So far, I’ve wired up and tested the basic motor system using L298N motor drivers, and I’ve worked out the control logic so that each wheel responds correctly. One of the motors (rear right) was slower than the others, so I had to compensate for that in the code by running it at a higher duty cycle.
The brain of the bot is where things get fun. Glide X uses a YDLIDAR T-mini Plus for mapping its surroundings and enabling autonomous movement. I’m setting it up for SLAM (Simultaneous Localization and Mapping) so that it can explore, understand, and navigate a warehouse-like environment.
Later, once the Pi Camera Module 3 (NoIR Wide) arrives, I’ll be integrating YOLOv8 for object detection. That way, Glide X can visually verify if the right product has been picked up before loading.
Apart from just moving around, Glide X will also have an actuator-powered truck bed to lift and lower packages, and a 5-DOF robotic arm to pick, place, or sort items once they’re verified by the camera. This combination allows Glide X to actually handle physical tasks, not just move from point A to B.
The entire system is going to be controlled from a custom web dashboard that I’ll build using Flask and WebSockets. From this dashboard, I’ll be able to:
In addition to remote control, I’ve added an inbuilt touchscreen mounted directly onto the robot. This allows Glide X to work in a kiosk-like mode, where local staff can interact with it directly — whether to trigger a dispatch, view delivery status, or run a quick diagnostic.
Done So Far:
Coming Up:
This has been one of the most rewarding builds I’ve worked on. It’s still evolving, but every part I finish brings it closer to being a fully autonomous, AI-powered warehouse assistant. I’ll keep updating this space as more modules go live and Glide X starts rolling into real tests!