No items found.
Robotic Inverse Kinematics Engine
Containerized inverse kinematics service for robotic joints in simulated space, offering HTTP API for joint states, translation paths, and collision avoidance. Enables real robotic hardware execution.
Client
Advanced Solutions Life Sciences
Date
June 1, 2020
Service
Software Engineering
Location
Louisville - KY (Remote)
Challenges
- Complex Algorithms: Developing accurate inverse kinematics algorithms for diverse robotic structures can be challenging.
- Simulation Accuracy: Simulated spaces may not perfectly mirror real-world scenarios, leading to disparities in solutions.
- Collision Detection: Accurate collision avoidance algorithms are crucial to prevent damage and ensure safety.
- Real-time Execution: Achieving real-time execution of computed joint states on hardware demands efficient coding and communication protocols.
- Scalability: Handling a large number of requests and diverse robotic systems while maintaining performance can be complex.
- Integration Issues: Integrating the solution with various hardware setups and software architectures might pose compatibility problems.
- Data Security: Ensuring the security of data transmitted through the HTTP API, especially if it involves sensitive information.
- Resource Optimization: Efficiently managing computational resources within the containerized environment for optimal performance.
- Continuous Testing: Rigorous testing to validate the solution's accuracy, especially when translating simulation results to real-world actions.
- Maintenance: Regular updates and maintenance to adapt to evolving robotic technologies and user requirements.
Project Results
The project successfully achieved significant milestones in developing a robust solution for complex robotic joint movements. Beginning with the extraction of precise robotic body parts from CAD models, the team meticulously optimized them using Autodesk Inventor and Blender 3D. Leveraging ROS, MoveIT, and the URDF standard, they constructed a virtual representation of the proprietary robotic hardware, meticulously validated in the Gazebo visualization tool. Complex pathing algorithms were applied, and planning was executed using MoveIT and RViz, ensuring the accuracy of joint movements. Real-time collision avoidance was integrated using Intel RealSenseTM Depth Camera SR305, enhancing safety protocols. The team ingeniously employed ROS Publish/Subscribe architecture to build a Python HTTP API, bridging the virtual and real robotic worlds seamlessly. Additionally, the implementation of a multi-step joint trajectory computation paved the way for replication in real hardware. Through these tasks, the project not only tackled the challenges but also delivered a sophisticated, integrated solution that promises advancements in robotic movement, setting a benchmark for future endeavors in this domain.