Publications

Autonomous manipulation is a challenging problem in field robotics due to uncertainty in object properties, constraints, and coupling phenomenon with robot control systems. Humans learn motion primitives over time to effectively interact with the environment. We postulate that autonomous manipulation can be enabled by basic sets of motion primitives as well, but do not necessitate mimicking human motion primitives. This work presents an approach to generalized optimal motion primitives using physics-informed neural networks. Our simulated and experimental results demonstrate that optimality is notionally maintained where the mean maximum observed final position percent error was 0.564% and the average mean error for all the trajectories was 1.53%. These results indicate that notional generalization is attained using a physics-informed neural network approach that enables near optimal real-time adaptation of primitive motion profiles.

Abstract not provided.

Autonomous and semi-autonomous robot manipulation systems require fast classification and localization of objects in the world to realize online generation of motion plans and manipulation waypoints in real-time. Furthermore, constraints and estimated plausible motions of objects of interest in space is paramount for autonomous manipulation tasks. For nongrasping tasks like pushing a box or opening an unlatched door, physical properties such as the center of mass and location of constraints like hinges or bearings must be considered. This paper presents a methodology for rapidly inferring constraints and motion plans for objects of interest to be manipulated. This approach is based on a combination of object detection, instance segmentation, localization methods, and algebraically relating different semantically labeled objects. These methods for motion estimation are implemented on a color-depth camera (RGB-D) and a 7 degree-of-freedom serial robot arm. The algorithm's performance is evaluated through different arm poses, assessing both centroid accuracy and estimation speed, and motion estimation performance. Algorithms are tested on an exemplar problem consisting of a block constrained on a dual linear rail system, i.e., constrained linear motion. Experimental results showcase the scalability of this approach to multiple classes with sublinear slowdowns and linear motion plan direction errors as low as 1.23E-4 [rad]. The manuscript also outlines how these methods for rapid constrained object motion estimation can be leveraged for other applications.

We study the problem of decentralized classification conducted over a network of mobile sensors. We model the multiagent classification task as a hypothesis testing problem where each sensor has to almost surely find the true hypothesis from a finite set of candidate hypotheses. Each sensor makes noisy local observations and can also share information on their observations with other mobile sensors in communication range. In order to address the state-space explosion in the multiagent system, we propose a decentralized synthesis procedure that guarantees that each sensor will almost surely converge to the true hypothesis even in the presence of faulty or malicious agents. Additionally, we employ a contract-based synthesis approach that produces trajectories designed to empirically increase information-sharing between mobile sensors in order to converge faster to the true hypothesis. We implement and test the approach on experiments with both physical and simulated hardware to showcase the approach's scalability and viability in real-world systems. Finally, we run a Gazebo/ROS simulated experiment with 12 agents to demonstrate the scalability of our approach in large environments with many agents.

Abstract not provided.

Abstract not provided.