Robot Configuration - For the Open Platform League (OPL)
Embedded Files
Robocup@Home 2024 Eindhoven
Meet our advanced service robot tailored for the RoboCup@Home competition!
Robot base with battery, sensors and computing units (Laptop, Jetson Xavier, RPi 4)
3x Omni wheel reconfiguration with custom actuator units driven using ODrives
6-dof Robot arm (XArm 6) for manipulating objects
3 Fingered soft gripper for robust yet simple grasping capabilities
Lidar (Livox MID-360) for SLAM and autonomous navigation
Zed Mini stereo camera for object detection
The video to the right is the current culmination of technical implementations!
Current configuration of components on the robot.
Three computing units for low-level control, central computing, and visual perception
Onboard power electronics
Sensors and other interfaces to interact with the external world
Manipulation and Grasping
Manipulation is a crucial function for service robots, as it involves the ability to effectively handle and interact with objects in various settings. This capability allows robots to grasp a wide range of objects, from small and delicate items to larger, heavier ones, and place them accurately in different locations within a home environment. For instance, the robot needs to organize items in cabinets, load and unload dishwashers, or even open and close doors. A high level of proficiency in object manipulation and grasping not only enhances the robot's utility but also makes it more versatile and adaptable to a variety of household tasks.
Computer Vision
Perception is a crucial aspect of service robotics, as it allows robots to understand and interact with their environment through sensors such as cameras. Using computer vision, robots needs to detect and classify objects, recognize and remember individuals and faces, and interpret gestures. By accurately perceiving their surroundings, robot can operate autonomously, make informed decisions in real time, and effectively assist with tasks such as guiding people, delivering items, or performing other services.
Object Detection
Human Pose Estimation
Autonomous navigation
The capability to create a map, detect and avoid obstacles, and accurately localize itself is crucial for service robots. These functionalities enable the robot to navigate effectively in various environments. Accurate mapping allows the robot to understand its surroundings and plan efficient routes. Obstacle detection and avoidance ensure the robot can safely maneuver around objects or people, preventing collisions. Localization, on the other hand, helps the robot identify its position within the environment, which is essential when the robot is tasked with reaching a specific location or performing precise actions. Together, these abilities are fundamental to the robot’s overall performance, making it capable of executing tasks reliably and autonomously.
Navigation test in simulation
Navigation test in real life. Static obstacles.
Navigation test in real life. Dynamic obstacles.
Human Robot Interaction and Natural Language Processing
The interaction between a robot and a human is crucial for the effectiveness of a service robot. The robot must be able to communicate effectively, understanding human commands and needs, and providing appropriate responses. Based on the human's requirements, the robot should then perform the necessary actions or assist the person in completing tasks, such as retrieving items, helping with household chores, or providing assistance.
Planning
Finally, integrating all the aforementioned functions, such as object manipulation, grasping, and human-robot interaction, and then planning the robot's actions to decide what to do and when to do it is essential for overall performance, especially in competitive settings. Effective planning enables the robot to determine the appropriate actions in various scenarios, adapting its behavior based on changing conditions and tasks. A behavior tree is often employed for this purpose, as it provides a structured and flexible framework for decision-making and action planning. This approach allows the robot to execute complex sequences of actions, manage priorities, and respond dynamically to different situations, ensuring that it performs tasks efficiently and accurately during competitions or real-world applications.
Page updated
Google Sites
Report abuse