... Advanced Search. Look in. Subject Author. Title Abstract. Date Range. Recent Advances in Dist... more ... Advanced Search. Look in. Subject Author. Title Abstract. Date Range. Recent Advances in Distributed Tactical Surveillance. Mahesh Saptharishi, Kiran Bhat, Chris Diehl, Spence Oliver, Marios Savvides, Alvaro Soto, John Dolan, and.
Decentralized multirobot learning refers to the use of multiple learning entities to achieve the ... more Decentralized multirobot learning refers to the use of multiple learning entities to achieve the optimal solution for the overall robot system. We demonstrate that single-robot learning theory can be successfully used with multirobot systems, but with certain conditions. The success and the effectiveness of this method are potentially affected by various factors that we classify into two groups: the nature of the robots and the nature of the learning entities. Incorrect setup of these factors may lead to undesirable results. In this paper, we methodically test the effect of varying four common factors (reward scope, learning algorithms, diversity of robots, and number of robots) in a decentralized multirobot system, first in simulation and then on real robots. The results show that two of these factors, reward scope and learning algorithm, if set up incorrectly, can prevent optimal, cooperative olutions. s 1.
Filament winding of fiber-reinforced composites requires precise control of the filament band pat... more Filament winding of fiber-reinforced composites requires precise control of the filament band path on the mandrel surface onto which it is wound. In helical filament winding, a basic path is selected for traversing the mandrel once. This path is then shifted an appropriate number of times in order to obtain a closed or near-closed filament layer. The current paper corrects a flaw in previous presentations [1, 2, 3] of an axisymmetric closure algorithm. Closure algorithms and examples of their implementation are also presented for smooth and polygonal non-axisymmetric parts of constant cross-section ("cylinders"). These algorithms can be used to wind non-axisymmetric cylindrical parts on standard two-degree-of-freedom winding machines.
This paper focuses on the development of an advanced telesupervision system architecture that sup... more This paper focuses on the development of an advanced telesupervision system architecture that supports a highly efficient approach to human-robot interaction while allowing heterogeneous robotic assets to be deployed. The described multi-level framework not only supports human monitoring and control of existing state-of-the-art autonomous agents, but also allows new autonomous agents to be incorporated at each level as they become available. Our philosophy of maximizing the efficiency and safety of humans through telesupervision of autonomous robotic systems is applicable in extreme environments on Earth, and across space including telesupervising Lunar and Martian robots when humans are nearby. We describe two applications of our Multilevel-Autonomy Robot Telesupervision Architecture: for planetary mineralprospecting using multiple semi-autonomous rovers conducted under a past project funded by the NASA Exploration Systems Mission Directorate; and for Harmful Algal Bloom detection ...
Proceedings of the AAAI Conference on Artificial Intelligence
Control Barrier Functions have been extensively studied to ensure guaranteed safety during inter-... more Control Barrier Functions have been extensively studied to ensure guaranteed safety during inter-robot interactions. In this paper, we introduce the Parametric-Control Barrier Function (Parametric-CBF), a novel variant of the traditional Control Barrier Function to extend its expressivity in describing different safe behaviors among heterogeneous robots. A parametric-CBF based framework is presented to enable the ego robot to model the neighboring robots behavior and further improve the coordination efficiency during interaction while enjoying formally provable safety guarantees. We demonstrate the usage of Parametric-CBF in behavior prediction and adaptive safe control in the ramp merging scenario.
... Advanced Search. Look in. Subject Author. Title Abstract. Date Range. Recent Advances in Dist... more ... Advanced Search. Look in. Subject Author. Title Abstract. Date Range. Recent Advances in Distributed Tactical Surveillance. Mahesh Saptharishi, Kiran Bhat, Chris Diehl, Spence Oliver, Marios Savvides, Alvaro Soto, John Dolan, and.
Decentralized multirobot learning refers to the use of multiple learning entities to achieve the ... more Decentralized multirobot learning refers to the use of multiple learning entities to achieve the optimal solution for the overall robot system. We demonstrate that single-robot learning theory can be successfully used with multirobot systems, but with certain conditions. The success and the effectiveness of this method are potentially affected by various factors that we classify into two groups: the nature of the robots and the nature of the learning entities. Incorrect setup of these factors may lead to undesirable results. In this paper, we methodically test the effect of varying four common factors (reward scope, learning algorithms, diversity of robots, and number of robots) in a decentralized multirobot system, first in simulation and then on real robots. The results show that two of these factors, reward scope and learning algorithm, if set up incorrectly, can prevent optimal, cooperative olutions. s 1.
Filament winding of fiber-reinforced composites requires precise control of the filament band pat... more Filament winding of fiber-reinforced composites requires precise control of the filament band path on the mandrel surface onto which it is wound. In helical filament winding, a basic path is selected for traversing the mandrel once. This path is then shifted an appropriate number of times in order to obtain a closed or near-closed filament layer. The current paper corrects a flaw in previous presentations [1, 2, 3] of an axisymmetric closure algorithm. Closure algorithms and examples of their implementation are also presented for smooth and polygonal non-axisymmetric parts of constant cross-section ("cylinders"). These algorithms can be used to wind non-axisymmetric cylindrical parts on standard two-degree-of-freedom winding machines.
This paper focuses on the development of an advanced telesupervision system architecture that sup... more This paper focuses on the development of an advanced telesupervision system architecture that supports a highly efficient approach to human-robot interaction while allowing heterogeneous robotic assets to be deployed. The described multi-level framework not only supports human monitoring and control of existing state-of-the-art autonomous agents, but also allows new autonomous agents to be incorporated at each level as they become available. Our philosophy of maximizing the efficiency and safety of humans through telesupervision of autonomous robotic systems is applicable in extreme environments on Earth, and across space including telesupervising Lunar and Martian robots when humans are nearby. We describe two applications of our Multilevel-Autonomy Robot Telesupervision Architecture: for planetary mineralprospecting using multiple semi-autonomous rovers conducted under a past project funded by the NASA Exploration Systems Mission Directorate; and for Harmful Algal Bloom detection ...
Proceedings of the AAAI Conference on Artificial Intelligence
Control Barrier Functions have been extensively studied to ensure guaranteed safety during inter-... more Control Barrier Functions have been extensively studied to ensure guaranteed safety during inter-robot interactions. In this paper, we introduce the Parametric-Control Barrier Function (Parametric-CBF), a novel variant of the traditional Control Barrier Function to extend its expressivity in describing different safe behaviors among heterogeneous robots. A parametric-CBF based framework is presented to enable the ego robot to model the neighboring robots behavior and further improve the coordination efficiency during interaction while enjoying formally provable safety guarantees. We demonstrate the usage of Parametric-CBF in behavior prediction and adaptive safe control in the ramp merging scenario.
Beatrice Institute. Pittsburgh [Remote], Jan 25, 2021
A conversation between theologian Jordan Wales (Hillsdale College) and scientist John Dolan (Robo... more A conversation between theologian Jordan Wales (Hillsdale College) and scientist John Dolan (Robotics Institute, Carnegie Mellon) on the theological and ethical implications of interacting with artificial intelligence, particularly androids. How will we relate to androids? Will we end up treating androids as people, allowing them to become substitutes for genuine relationships? And how will our interactions with androids shape our relationships with each other? Will they form us to dehumanize actual persons? Dr. Wales and Dr. Dolan will address the science of artificial intelligence, what robots can currently and will soon be able to do, as well as the ways in which humans relate to artificial intelligence in order to help us conceive of a proper place for artificial intelligence in a flourishing society.
Uploads
Recording: https://www.youtube.com/watch?v=TRjvNqi8DdQ