This paper presents a prototype Automatic Guided Vehicle (AGV) equipped with two types of proximity sensors. Although largely modeled as a line-following robot, the algorithm makes reactive decisions based on sensory data to avoid the obstacles by temporaily deviating from the predefined path. The infrared (IR) sensors recognize the desired path of the AGV by detecting the high-contrast line on the surface, thus enabling automated motion of the AGV. The ultrasonic proximity sensors on the AGV aid in the detection of the obstacles or for following moving objects such as another AGV. Different scenarios are discussed where this multisensor capability enables enhanced perception for AGVs without resorting to a computationally more intensive robot vision system

This work introduces a digital-twin based approach to path planning in the presence of obstacles for a quadrupedal robot, utilizing virtual sensors and a path planning algorithm. Ray-casting and NavMesh agent-based algorithms are used to devise collision-free paths. The Ray-casting algorithm detects dynamic objects virtually, while the NavMesh system determines a feasible path for the robot. Real-time communication between the digital twin and the physical robot facilitates seamless movement. The effectiveness of this approach in enhancing the robot’s navigation and obstacle avoidance capabilities can be demonstrated through both simulated scenarios and real-world experiments. The implementation of the digital twin, along with the utilization of path planning and obstacle avoidance algorithms available in the Unity software, as well as the real-time integration with the physical robot, are significant contributions of this work. A shared environment is demonstrated for communications between robots, each with its own digital twin

Constricted passages such as tubes may be out of the view of the robot vision system, and the wrist-force data may be useful in such cases. Collaborative robots (Cobots), which are equipped with precision wrist-force sensors, are suitable for sharing the workspace safely with human workers and other robots and devices. The data from the force sensor allows building more intelligent algorithms for robot control, and plan access to relatively unstructured obstacles. In this work, an attempt is made to model obstacles and constricted passages graphically for the digital-twin implementation of a collaborative robot and use the same for motion planning in the absence of vision sensors. The main scope of this work is to build the graphical framework for the digital twin and link the same with path planning algorithms in the public domain using MoveIt and Robot Operating System (ROS). Motion visualization is implemented using another software, called RViz. The co-simulation between MoveIt and RViz is, in turn, linked to the physical robot, or optionally, the simulation environment provided by the robot supplier

Collaborative Robots (Cobots) constitute a new class of industrial robot manipulators that are now becoming popular in tasks that require intelligent manipulation and human–robot interaction. The characteristic difference between these robots and the conventional industrial robots is the availability of force feedback from the wrist-force sensor. This allows the robot arm to avoid obstacles, detect variation in the part locations and estimate the type of objects being handled based on the weight or contact forces. Digital twin is the representation of a real-world object in a virtual environment. This allows the synchronization and replication of the motions of the robot present in the real world with that of the counterpart in a virtual environment and vice versa. This work is an initial attempt to develop a remote interface and a digital twin for a collaborative robot (Universal robot UR5e). Interfacing of the robot through Robotics Operating System (ROS) and the use of Unity3D for the development of the digital twin are reported here.