Leo Rover: Autonomous Mobile Robot
A physical Leo Rover built up and deployed by a five-person team: SLAM, path planning and real-time obstacle detection with ROS 2, tested in the lab arena.
Problem to result
Who had the problem, what I built, the hardest part, and what it achieved.
A five-person team had to get a Leo Rover to explore an unknown arena on its own, find coloured cubes and pick them up in a single 20-minute mission, where one failure (a blocked sensor, a dropped link, a missed grasp) ended the run.
A rover that maps an unknown arena, spots coloured cubes with its camera and picks them up with its arm, with no one driving it.
Fitting an arm, gripper, onboard computer, camera and electronics on a small rover without blocking the LiDAR's 360° view or making it tip over. I designed an elevated two-plate sled that sits 4 mm above the LiDAR, mounted the arm front-centre and made the computer plate movable to balance the weight, then split power into two battery domains so the arm's current spikes couldn't destabilise the computer.
Navigation accuracy of 95%; a stability factor of 1.71 with the arm stowed and 1.58 extended; mounts with a safety factor of 2.248 at 1.5× load; about 1.4 h of battery against a 20-minute mission.
Key Features
Interactive 3D structural breakdown of the LEO Rover.
SLAM
Simultaneous Localisation and Mapping utilising 2D/3D LiDAR fusion to construct high-fidelity spatial maps of unknown environments in real-time.
Autonomous Navigation
Dynamic path planning and robust obstacle avoidance algorithms ensuring seamless point-to-point traversal across complex terrain.
Computer Vision
Edge-deployed YOLOv8 inference enabling rapid object detection, semantic segmentation, and advanced environmental perception.
ROS2 Architecture
A decentralised, highly modular communication framework managing sensor data streams and autonomous state machines continuously.
Path Logging
Comprehensive telemetry recording system tracking odometry, executed paths, and obstacle metadata for post-mission kinematic analysis.
Product Showcase
Deep dive into the specialized hardware components used in our mobile autonomous systems.

Hardware Module
RPLidar Sensor
A high-performance 360-degree laser range scanner. Provides the robot with precise spatial awareness and real-time mapping capabilities.
Design Files & CAD
Models, schematics, and assets.
Autonomous Mobile Robot Assembly
Live Simulation Output
Simulated console execution.
Project Walkthroughs
Click any video to play.
LivePerception System (Rover's PoV)
Object detection and LIDAR perception capabilities from the Rover's point of view.
3:48Arm Movement Testing
Testing the movement of the arm.
5:20Structural Construction
The construction of the robot.
7:15Trail Run
First run of the robot in a trail environment.
1:45Navigation Simulation
Autonomous Mobile Robot navigation simulation in a controlled environment.
2:15Autonomous Navigation
Autonomous Mobile Robot navigation simulation - Alternative scenario.
1:20Navigation toward to block
Autonomous Mobile Robot navigation toward a specific block.
LiveReturn to intial position after mission (Simulation Demo)
Simulation of the robot's navigation path.
LiveArm Manipulation (Simulation)
Simulation of the robot's arm manipulation.
LiveEmergency brake (Simulation)
Demonstration of the robot's safety features and protocols.
LiveFinal Evaluation
Testing the integrated SLAM, path planning, and YOLOv8 object detection systems during an autonomous run.
Problems & Fixes
What went wrong while building it, and how each problem was solved.
Five parts (arm, gripper, NUC, camera, electronics) all needed space, and a flat layout put them in the LiDAR's scan plane, blocking much of its 360° view and degrading SLAM
Built an elevated sled. The M5 × 50 mm spacers put the upper plates 4 mm above the LiDAR's top (41 mm LiDAR + 5 mm mount = 46 mm), so the 360° view stays clear
Get code
GitHub repositories for this project.
Leo Rover: Autonomous Mobile Robot Repository
Access the complete source code on GitHub.
Noticed by RealSense
This project caught the attention of RealSense's VP of Developer Ecosystem.
“Awesome! Great job Sarath Kumar Senthilkumar Kavitha and thanks for using RealSense!!”
Posted on teammate Sarath Kumar's write-up of the project. RealSense is the depth-camera platform behind the rover's perception pipeline (colour, shape and depth detection).
View the post on LinkedInTeam
The dedicated team of engineers and innovators behind the Autonomous Robot project.
Computer Vision and Navigation
Design, 3D printing and hand‑eye calibration
Navigation and SLAM
Navigation and spatial awareness
SLAM and spatial awareness








