You will get complete simulation RL Policy model with Isaac Sim/MuJoCo/Gazebo
Top Rated
Project details
My Recent Experience with Similar Projects
1. Dexterous Hand Grasping in MuJoCo with PPO
Objective: Simulate a multi-fingered robotic hand to grasp and lift various objects (cube, cylinder, bottle)
Tools: MuJoCo, Python, PPO (Stable-Baselines3), custom Gym environment
Work Done:
Loaded a URDF-based custom hand into MuJoCo
Designed object sets with randomized poses and physics parameters
Implemented joint-level position and torque control policies
Trained a PPO agent with custom reward shaping for grasp stability and object lift success
Verified robustness under external disturbances and domain noise
2. Robotic Arm & Hand Coordination in Isaac Sim
Objective: Train a 7-DOF arm with a 3-finger gripper to perform precise grasp-and-place tasks
Tools: Isaac Sim (Omniverse), PPO (SB3), Jetson Xavier NX, ROS bridge
Highlights:
Converted URDF to USD and scripted the full RL loop using Isaac’s simulation API
Used PPO with an LSTM policy to enable sequential decision-making
Integrated contact sensors and visual feedback for reward shaping
Achieved consistent grasp success across randomized object sizes and placements
3. RL-Based Grasp Strategy Optimization in PyBullet
1. Dexterous Hand Grasping in MuJoCo with PPO
Objective: Simulate a multi-fingered robotic hand to grasp and lift various objects (cube, cylinder, bottle)
Tools: MuJoCo, Python, PPO (Stable-Baselines3), custom Gym environment
Work Done:
Loaded a URDF-based custom hand into MuJoCo
Designed object sets with randomized poses and physics parameters
Implemented joint-level position and torque control policies
Trained a PPO agent with custom reward shaping for grasp stability and object lift success
Verified robustness under external disturbances and domain noise
2. Robotic Arm & Hand Coordination in Isaac Sim
Objective: Train a 7-DOF arm with a 3-finger gripper to perform precise grasp-and-place tasks
Tools: Isaac Sim (Omniverse), PPO (SB3), Jetson Xavier NX, ROS bridge
Highlights:
Converted URDF to USD and scripted the full RL loop using Isaac’s simulation API
Used PPO with an LSTM policy to enable sequential decision-making
Integrated contact sensors and visual feedback for reward shaping
Achieved consistent grasp success across randomized object sizes and placements
3. RL-Based Grasp Strategy Optimization in PyBullet
AI Development Type
Deep Learning, Model TuningAI Tools
Keras, Open Neural Network Exchange, OpenCV, PyBrain, PyTorch, TensorFlowAI Development Language
PythonWhat's included
| Service Tiers |
Starter
$250
|
Standard
$350
|
Advanced
$500
|
|---|---|---|---|
| Delivery Time | 3 days | 5 days | 8 days |
Number of Revisions | 2 | 4 | 6 |
AI Model Integration | |||
Detailed Code Comments | |||
Knowledge Graph | - | ||
Model Documentation | - | ||
Ontology | - | - | |
Source Code | - | - | |
Taxonomy | - | - |
38 reviews
(33)
(1)
(2)
(1)
(1)
This project doesn't have any reviews.
TG
Tuatini G.
Jun 19, 2026
Review a ROS 2 alternative framework
Avanik and his team were great! Communication was smooth and they were able to successfully deliver on the project we have them.
RL
Rhys L.
Apr 15, 2026
30 minute consultation
BY
Ben Y.
Mar 7, 2026
RL Expert Needed for Multi-SKU Inventory Optimization
DL
Daniel L.
Feb 11, 2026
Review a ROS 2 alternative framework
Avanik is a fantastic engineer, very responsive with very constructive feedback. Would recommend to anyone.
AB
Ashton B.
Jan 25, 2026
Looking for a professional Robotics Engineer
About Avanik
Robotics/Automation R&D Engineer | Solidworks | ROS | AI | ML/RL | LLM
91%
Job Success
Stuttgart, Germany - 11:44 pm local time
âś… I am a Robotics/Automation Engineer with 5+ years of professional experience in various fields. I am willing to travel internationally for work. I specialize in developing real-time motion control RL-based policies for humanoid robots, such as the Unitree G1, encompassing simulation, gait optimization, hardware deployment, and neural inference.
I am proficient in:
• Smart robotics engineering (from mechanical to software end-to-end solution)
• A passion for robots that solve problems or provide a meaningful purpose
• New concept R&D design and development of any SPM system, Automation Conveyor system, Humanoid robot, Robotic Arms, etc.
• Path planning and navigation planning system on ROS, Isaac Sim, MuJoCo simulation
• 3D modeling, 2D drawing, and Engineering Drawing using SolidWorks and Inventor
• 3D model rendering/animation utilizing Blender, Unity, Unreal Engine, and Keyshot
• ML / RL & AI algorithms, LLM model development, and also experts in RoboDK
• Highly experienced in working with actuators and motors in machinery development
• Brief evolution in Reverse engineering
• Have worked with many companies, from startups to corporate R&D groups
• I have an extensive network of industry specialists and vendor suppliers, both domestic and international. If it can be defined, it can be made
âś… As a blockchain developer, I led a team under my supervision. All of them are working nowadays and have many clients. We have worked mainly on services in Crypto NFT, Blockchain, Unity 3D, and Metaverse development.
âś… I am eager to work with you to meet your expectations and build a long-term relationship. Please get in touch with me to discuss your project.
Steps for completing your project
After purchasing the project, send requirements so Avanik can start the project.
Delivery time starts when Avanik receives requirements from you.
Avanik works on your project following the steps below.
Revisions may occur after the delivery date.
Share a brief project description
After purchasing the project, send the requirements so Adnan can start the project
Works on your project following the steps below
Ping me on Upwork