Study and Development of Strider Linkage Legged Robot's Movement with HuskyLens AI Vision Camera

Abstract

This paper presents the design, development and evaluation of a legged robot based on a 10-bar Strider linkage mechanism, aimed at achieving stable straight-line locomotion with and without a HuskyLens AI vision camera. Unlike conventional legged robots that depend on complex gait algorithms, this approach leverages mechanical linkage geometry for natural walking motion. The robot chassis and legs were fabricated using Polyethylene Terephthalate Glycol (PETG) via 3D printing due to its superior tensile strength (~50 MPa) and high heat resistance. An Arduino Uno microcontroller, integrated with a HuskyLens AI vision camera, was used for line tracking and control. Performance testing assessed time, speed and energy efficiency across 20 experimental trials over a 5-meter course. Quantitative results indicate that while the mechanical design alone allows straight movement initially, open-loop cumulative drift caused the unassisted robot to fail completely after Trial 10 (0% success rate from Trial 11 onwards). Conversely, the AI-enhanced control closed the feedback loop, ensuring a 100% trial completion rate while mitigating drift through real-time trajectory adjustments, demonstrating a robust trade-off between minimal processing overhead and path-tracking accuracy.