Human-State-Aware Non-Linear Control Framework for Physical Collaborative Aerial Transportation

Recent advancements in mechanical design and robust control modalities have significantly enhanced the ability of aerial robots to interact with humans in various physical tasks. A critical challenge remains to ensure safe, cooperative, and intuitive collaboration between humans and aerial robots during such interactions. In this context, this work introduces a human-state-aware nonlinear control framework specifically designed for collaborative object transportation involving a human operator and a quadcopter. A system is considered in which a rigid object is lifted by a human using a custom-built sensor device, the Human Handle Device. At the same time, the other end is suspended from the quadcopter using a cable. The proposed control modality is designed to improve the quadcopter’s tracking capabilities and is proven exponentially stable. The controller’s performance is justified by conducting experiments with and without feed forwarding the human states, particularly the human acceleration, in the quadcopter’s controller. The experimental results validate the proposed approach, demonstrating that incorporating human states allows the quadcopter to respond effectively to human actions. This leads to smoother and more stable tracking performance, enhancing overall cooperation and interaction in shared-object transportation tasks.