Grant: Discovery Grants Program - Individual

Robotic grasping stands at the frontier of engineering and technology because of its potential to revolutionize how automated systems interact with physical environments. The mastery of multi-finger grasping techniques opens new possibilities for advancements in several fields such as manufacturing, healthcare, and agriculture. However, despite these apparent benefits and the field's maturity, robotic grasping still faces substantial hurdles, primarily due to the complexity of replicating human dexterity and adaptability. This research aims to bridge the gap between mechanical action and human-like dexterity, paving the way for robots to perform complex manipulations, ultimately leading to societal progress and economic resilience. In this research, we concentrate on 3 objectives, each foundational to enhancing the capabilities of robotic systems. In Objective 1, we develop a new grasp-planning strategy for multi-finger robotic hands. Our approach uniquely combines the reliability of analytical methods with the strength of data-driven algorithms. Central to this integration is prioritizing thumb placement, an inspiration drawn from human grasping. We identify a general region for thumb placement using data-driven insights, then employ analytical precision to obtain an optimal multi-finger grasp. This novel approach addresses the challenges unmet by either method alone. In Objective 2, we address the necessity for robots to maintain stable grasps by incorporating variable stiffness models directly into the grasp force analysis, moving away from traditional rigid models. Our approach allows for more nuanced control strategies, bridging the gap in dynamic adaptability between robotic and human hands. In Objective 3, we acknowledge the fundamental role of hardware, especially actuators, in realizing the strategies developed in the first 2 objectives. Our goal is to engineer innovative miniaturized actuators with torque-control capabilities, building on our extensive research in magneto-rheological devices. These actuators will underpin the construction of multi-finger robotic hands with enhanced functionality. This research sets the trajectory for a future where robots can emulate the intricate coordination observed in humans when manipulating objects, and, much like humans, perform real-time force adaptations during grasp. The long-term impact of this research is profound, extending beyond technical gains. Success in this domain means safer working conditions, reduced labor costs, and increased efficiency in resource allocation, all converging towards a sustainable and technologically empowered future. Moreover, the program contributes to the training of an inclusive, diverse talent pool in a thriving knowledge ecosystem. The research, in addition to pushing the technological boundaries, will also drive societal transformation, ensuring that the progress made reflects a multitude of perspectives and is accessible to a broader spectrum of society.