Mohamed M.Private profile: Upwork shows its earnings only to signed-in members, so we check it every few weeks, not daily. MRR here is a rough estimate (Upwork's six-month earnings ÷ months active). Last read Nov 5, 2026.

I take robots from lab demos to real-world deployment. My day job is shipping commercial multi-robot fleets, and on Upwork I'm Top Rated Plus with 100% Job Success. I work on quadrupeds, humanoids, GPS-guided ground robots and drones. Two kinds of projects fit me best: robots that have to navigate reliably outside the lab, and humanoid or arm projects that need a working path from simulation and teleoperation data to a policy running on real hardware. WHAT I CAN BUILD FOR YOU Manipulation and humanoids - VR teleoperation for Unitree humanoids (G1, R1) with Meta Quest 3 and xr_teleoperate, set up so the same rig records demonstration data - Imitation learning and VLA pipelines: demonstration datasets, fine-tuning, and the policy deployed as a ROS 2 node on Jetson - VLA model selection for commercial products. I've evaluated GR00T N1.5 and UnifoLM-VLA for deployment, including the licence terms that decide whether you can ship them. Sim-to-real and real-to-sim - NVIDIA Isaac Sim and Gazebo environments for Unitree Go2, G1 and H1, ground robots and drones - Digital twins from real spaces. I rebuilt a real office as a 3D Gaussian Splat scene in Isaac Sim, with a Unitree H1 humanoid walking through it. - The same ROS 2 interfaces in simulation and on hardware, so the stack you test in sim is the stack that runs on the robot LLM and voice-controlled robots - Voice-to-action pipelines: speech recognition (Whisper), LLM reasoning, ROS 2 command execution. I built one for a Go2 in Isaac Sim and designed an Arabic voice assistant architecture for a G1 humanoid. Autonomous navigation - Full Nav2 stacks: GPS waypoint following, costmap tuning, elevation mapping, obstacle avoidance, multi-robot support SLAM and localization - LiDAR-inertial odometry (FAST-LIO2, GLIM, Point-LIO), robot_localization EKF, dual-antenna RTK GNSS (Septentrio, u-blox), NTRIP corrections. I've debugged RTK issues down to the raw SBF logs. Perception and edge AI - YOLO, TensorRT, NVIDIA DeepStream, LiDAR-camera calibration and fusion (Ouster, Livox, Hesai, RealSense), Jetson deployment Drones and UAVs - PX4 firmware modifications, custom obstacle avoidance, uORB messaging PLATFORMS I'VE SHIPPED ON Unitree Go2 and G1 humanoid, Husarion Panther, Clearpath Jackal, custom inspection robots, PX4 drones, AUVs. HOW I WORK Docker-based ROS 2 setups (Humble or Jazzy), documented handoffs with runbooks, honest estimates, and fluent English and Arabic. Tell me what your robot needs to do and where it needs to run. I'll tell you how I'd approach it and whether I'm the right fit.