Design and Development of Leo: An Affordable Biomechanically Inspired Quadruped Robot with Cognitive Abilities

This paper presents the design, development, and testing of Leo, an affordable biomechanically inspired quadruped robot with cognitive abilities. The objective was to create an accessible, low-cost platform for robotics research and education. Leo features a lightweight 3D-printed structure weighing 2 kg, with dimensions of 550 mm x 400 mm x 100 mm. The robot employs a novel crankshaft mechanism in its leg design to reduce torque requirements and improve efficiency. Leo's control system integrates Robot Operating System (ROS) and Arduino, allowing for both manual and semi-autonomous operation. Cognitive abilities are demonstrated through voice recognition, image recognition, and object detection capabilities using YOLOv4 and face recognition algorithms. Gait analysis reveals an elliptical trajectory with each walking cycle lasting 19–20 seconds. Real-world testing showed a close correlation between planned and actual movements, with an approximate error of 25% due to practical conditions. Leo represents a significant step towards creating accessible quadruped robots for research, household use, and small-scale tasks, with potential applications in education, home assistance, and prototype testing for more advanced robotics systems.