Find Robot Operating System Projects

This project involves modifying the calf (lower-leg) structure in the Unitree Go2 Pro’s 3D model. Existing 3D models of the Unitree Go2 Pro can be found online and are also available on Unitree’s GitHub.

In this project, we will redesign the calf section so that, when fully extended, the calf and thigh align in a straight line. This modification will be applied to all four legs.

Based on the reference diagrams, the goal is to create a SolidWorks assembly file in which each joint can be moved and tested to verify the new kinematic structure.

Development of control, ROS2 integration, and Gazebo simulation will proceed separately after the modified 3D model is completed.
In the Unitree Go2 Pro quadruped robot, the positions of the hip, thigh, and calf motors and gears will remain unchanged, but the structure of the calf section will be modified. The calf legs of all four limbs will be redesigned so that, when fully extended as shown in the diagram, they form almost a straight line. The length of the ground-contacting portion of the leg will remain nearly the same as the current design, and the hinge will be moved as far outward as possible so that, when pushed by the motor, it aligns almost linearly with the thigh.
The front legs will maintain the current knee joint orientation and bending direction, but the rear legs will be modified to bend backward like a human knee. In other words, by swapping the outer casing of the thigh and calf leg on the left and right sides of the rear legs, the current inward-bending geometry will be reversed so that the legs bend outward instead.
More specifically, the structure involves swapping the left and right orientation of the rear leg’s thigh and calf motor assembly, along with the lower leg structure, at the existing hip joint.

Skills: 3D Rendering, Solidworks, Mechanical Engineering, 3D Modelling, 3D Animation, Robotics, 3D Design, 3D CAD, Robot Operating System (ROS), Gazebo

I am building a ground robot that must move reliably both inside a warehouse and outside across uneven pavement, all under ROS 2 (Humble/Foxy—happy to align on version). I already have the mechanical platform; what I still need is a solid navigation-and-control layer that lets the robot localize itself, generate safe paths, and execute them smoothly.

Your task is to design, implement, and test the full ROS 2 navigation stack for this vehicle. That includes writing and tuning the nodes for localization, global + local planning, and velocity control, wiring them to my existing wheel-odometry, IMU, and LiDAR topics, and delivering the launch files, parameter sets, and concise documentation so I can reproduce the results on my hardware.

To sign off the work, I will run the supplied launch files on the physical robot in two scenarios—an indoor maze and an outdoor yard—expecting the robot to reach user-defined waypoints without collisions while maintaining reasonable speed limits. Provide any calibration steps I should follow and notes on further tuning so I can extend the system after delivery.

Skills: Python, C++ Programming, Robotics, Artificial Intelligence, Embedded Systems, Simulation, Control Engineering, Robot Operating System (ROS)

Hybrid fast DC charging system for electric vehicle (EV) batteries, combining photovoltaic (PV) energy and the utility grid within a unified power architecture. The system is designed to deliver high charging power (50–100 kW) while ensuring operational reliability, energy efficiency, and seamless charging continuity under variable solar and grid conditions.
The proposed architecture is centered around a high-voltage DC bus, typically regulated at 750–800 V, which serves as the common coupling point between all energy sources and the EV charging module. This configuration allows flexible power flow control, simplified source coordination, and efficient integration of renewable energy into fast-charging infrastructures.
The primary objective of the system is to maximize solar energy utilization while guaranteeing the required charging power through intelligent support from the utility grid whenever photovoltaic production is insufficient.

Se debe desarrollar un programa completo en PolyScope para un robot Universal Robots, siguiendo las especificaciones indicadas en el PDF adjunto.

La práctica consiste en realizar un proceso de paletizado de 18 cajas, distribuidas en tres filas y dos columnas por nivel, teniendo en cuenta las dimensiones y pesos indicados en el documento de la practica robot
El programador debe:
– Configurar la herramienta de ventosa con las medidas y peso indicados.
– Configurar el payload correspondiente a caja + ventosa.
– Definir todos los puntos y movimientos necesarios para completar el paletizado.
– Activar y desactivar la ventosa mediante tool_out[0], tal como requiere el ejercicio.
– Incluir comentarios, avisos y organización por carpetas según los criterios del módulo.
– Entregar el proyecto en un archivo comprimido con todos los archivos del robot, incluido el archivo de instalación, listo para abrir en máquina virtual PolyScope.

No es necesario implementar lógica avanzada ni sensores externos; únicamente deben estar creados y funcionales los movimientos del robot y la estructura del programa.

Skills: PHP, C Programming, Software Architecture, C++ Programming, Robotics, Robotic Process Automation, Automation, Robot Operating System (ROS)

Hi,

I’m currently building an MVP of a lightweight, modular Ground Control Station (GCS) for drones, focused on real-time telemetry, mission planning, video, and especially advanced log analysis. The project is local-first, offline-capable, with a Node.js backend, Angular frontend, and an infrastructure based on Docker.

I’ll need support mainly on:

Flight log processing, analysis, and visualization

Simulator integration (SITL, replay, test scenarios)

Data robustness and reliability

For this, experience with ArduPilot and/or PX4 is very important, along with solid full-stack skills (JS/TS, frontend & backend).

Do you also have experience with web development and Docker/containerized environments?

If so, I’d be happy to have a quick call to see if it’s a good match.

Thanks!

I already have LunarSim running under ROS 2 Humble on Ubuntu 22.04.5 LTS and now need a plug-and-play execution layer that lets the simulated rover travel a user-defined distance in both X and Y while it autonomously skirts any rocks in its path.

Key points of the job
• Use the depth images published on /lunarsim/depth_image/depth as the sole sensing source.
• Apply a classical computer-vision approach (OpenCV, PCL or similar) to detect rocks—no machine-learning training loops, just reliable depth-based segmentation.
• When a rock is detected, perform a simple avoidance manoeuvre: veer around the obstacle, re-establish the original heading, and keep going. No global mapping or heavy path planning required.
• Drive the rover exclusively through velocity commands so it behaves smoothly inside the simulator.

Deliverables
• A clean ROS 2 Humble package containing the navigation node(s), publishers/subscribers, and any helper libraries.
• Launch file(s) that start the simulator interface, the obstacle-detection node, and the motion controller in one command.
• Parameters for target X/Y distance, speed, safety margin, and timeout, all exposed through ROS 2 params.
• README with build instructions (colcon), topic diagram, and a quick demo sequence I can reproduce.

Acceptance test
1. Build and source the package in a fresh workspace.
2. Launch the system; specify e.g., X = 10 m, Y = 5 m.
3. Rover reaches the requested offset within a reasonable tolerance (<10 cm) without colliding with any rocks, issuing only velocity commands throughout the run.

If something in my environment is unclear, let me know early—otherwise I look forward to integrating your package and watching the rover drive itself.

Skills: Python, CUDA, Machine Learning (ML), C++ Programming, Ubuntu, OpenCV, Computer Vision, Simulation, Robot Operating System (ROS)

I’m looking for a mechatronics or hardware engineer who can help build a working prototype of an automatic haircut machine.

What I need:
• Concept and mechanism design
• 3D CAD model
• Motor and electronics system
• Working prototype

Prefer someone with experience in robotics, motors, sensors, or consumer devices.

Timeline and cost can be discussed.

Goal of the project:
Set up a usable manipulation stack in a ROS2 + simulator-based environment.
Understand the performance bottlenecks of these manipulation stacks using existing ROS2 profiling tools.
(manually) for each stack, define what end-to-end objectives matter and try to specify relative importance of each path.

Steps
Identify a manipulation stack. Criteria:
must run on ROS2
must run on Gazebo-Sim / Isaac Sim (Gazebo is preferable since Isaac is a little more painful to setup)
Candidates:
https://moveit.ai/install-moveit2/binary/
https://developer.nvidia.com/isaac/manipulator

Use ros2_tracing: Tracing tools for ROS 2 to profile the system
What is ROS2 tracing? TL;DR: super light weight tracing tool based on LTTng v2.13 that can be used even in realtime applications with very low overhead. This dumps traces to files, which can then be analyzed by a bunch of tools.
Downstream stuff you can use to process traces:
ros-tracing/tracetools_analysis: default tool that ships with ROS2
skoudmar/Ros2TraceAnalyzer: This claims to be 30x faster than the default tool. Also, they seem to have a way to visualize all this information in a graph.
ROS 2 message flow analysis experiments
broadly looks like there are many tools (e.g. see this discussion thread: ROS 2 Visualization Tools for Architecture)
Write a report on findings and summarize. Furthermore document every single command and approach on how you got there starting from installing ROS2 all the day to project completion.

Skills: Technical Writing, Software Development, Documentation, Data Analysis, Simulation, Robot Operating System (ROS), Software Engineering, Gazebo

Hello,

I am looking for an experienced drone engineer to perform a complete ArduPilot programming/tune of my 17-inch quadcopter.

The quadcopter has the following hardware assembled:
TBS Source X frame, MatekH743 Slim, MatekM10Q GPS, motors 4215-320kv, TBS Lucid 12S 90A ESC, DJI O4 Pro, ELRS 2.4GHz, no battery.

The goal is to perform full autonomous missions reliably using QGC, with the longest possible one-way flight (30km ideally) and as much payload-carrying capability as possible.