Abstract:
In the era of Industry 4.0, leading industries are adopting sustainable solutions in their workplace. Industrial inspection is one such task that requires high precision and can be hazardous in compromised areas. Unmanned robots are being presented as a solution to this problem and they are performing well above the mark. In this regard, a project of an autonomous wheeled quadruped robot is presented that is a cost-effective design for a convertible wheeled to-legged robot that can adapt to any environment and can easily navigate autonomously. It has vision capabilities on-board which can be accessed remotely and can be used for object detection, tracking, and identification. It can be used in hazardous environments and can be equipped with any sensor. Its software solution is distributed as ROS packages making it highly modular, easy to customize, and upgradeable. The robot is designed using 3D printing technology. It uses vision and ranging sensors to generate area maps and localizes itself in any mapped or unmapped environment. Upon giving a destination, it can generate an optimal path plan. This plan is utilized to navigate the robot to its destination. We are using FPGA as the robot low level control. Using FPGA for motion control provides parallel computation of gait motion for synchronized joint response. Furthermore, the high frequency adjustable clock of FPGA reduces the response time resulting in swift action. The robot can use its vision sensors to detect, identify and track objects in real time. Furthermore, it has the capability to detect faces on custom trained dataset and can scan QR codes. It can be navigated remotely using remote server configurations.
