Humanoid Robo
Budget: $250 – $750 USD
I have already built the core frame, drivetrain and basic ROS nodes for a humanoid robot; now I need a robotics specialist to take it across the finish line so it can reliably handle industrial-grade assembly work on a factory line.
What’s left to do spans both hardware and software. On the mechanical side, wrist and gripper modules must be refined for repeatable pick-and-place, the torso needs final cable routing, and the protective covers should be redesigned for faster maintenance access. Electrically, I still require clean integration of the force-torque sensors, proximity safety scanners and the 48 V power distribution board.
The bigger challenge is the control stack:
• Extend the existing ROS 2 framework with MoveIt motion planning for dual-arm manipulation.
• Write C++ / Python nodes for vision-guided part recognition (OpenCV or equivalent) and real-time torque compliance.
• Implement error-recovery behaviours so the robot can detect a mis-pick and automatically retry without human intervention.
• Create a simulation in Gazebo (or your preferred tool) that mirrors the final kinematics, then validate paths before flashing the firmware.
• Provide clear documentation and a short video demo of the robot assembling a sample jig.
I am open to component or library suggestions as long as the final system remains ROS-compatible and serviceable with readily available parts. Safety standards such as ISO or ANSI can be discussed if you have prior experience certifying collaborative robots, but they are not a hard requirement at this stage.
If you have shipped similar assembly robots, send me a brief outline of your approach and estimated timeline, and we can dive straight into the technical details.
What’s left to do spans both hardware and software. On the mechanical side, wrist and gripper modules must be refined for repeatable pick-and-place, the torso needs final cable routing, and the protective covers should be redesigned for faster maintenance access. Electrically, I still require clean integration of the force-torque sensors, proximity safety scanners and the 48 V power distribution board.
The bigger challenge is the control stack:
• Extend the existing ROS 2 framework with MoveIt motion planning for dual-arm manipulation.
• Write C++ / Python nodes for vision-guided part recognition (OpenCV or equivalent) and real-time torque compliance.
• Implement error-recovery behaviours so the robot can detect a mis-pick and automatically retry without human intervention.
• Create a simulation in Gazebo (or your preferred tool) that mirrors the final kinematics, then validate paths before flashing the firmware.
• Provide clear documentation and a short video demo of the robot assembling a sample jig.
I am open to component or library suggestions as long as the final system remains ROS-compatible and serviceable with readily available parts. Safety standards such as ISO or ANSI can be discussed if you have prior experience certifying collaborative robots, but they are not a hard requirement at this stage.
If you have shipped similar assembly robots, send me a brief outline of your approach and estimated timeline, and we can dive straight into the technical details.
Related categories:
C Programming
Python
CUDA
C++ Programming
Robotics
OpenCV
Robot Operating System (ROS)
Gazebo