Penn1: Experimental Testbed on Wireless Autonomous Systems
Abstract
Penn1: Experimental Testbed on Wireless Autonomous SystemsPublicly Releasable Project SummarySubmitted to Dr. Behzad Kamgar-Parsi at ONRThe falling price/performance of sensors, processors, storage devices, and communications hardware in the last decade has led to a hyperconvergence framework where computation, storage, networking, and virtualization is tightly integrated into a software-centric architecture to increase scalability and reduce complexity. Coupled with advances in wireless technologies, progress in autonomous systems, and breakthroughs in machine learning and artificial intelligence, this exponential growth in technology allows us to rethink the concept of operations in complex and contested environments. Heterogeneous teams of intelligent systems consisting of autonomous, mobile agents can greatly extend the capabilities and the reach of soldiers, in yet-to-be-imagined ways. They can provide situational awareness and create a protective bubble around the critical human and infrastructure assets against dynamic threats. Such heterogeneous teams, including humans, can achieve unprecedented operational effectiveness and superiority through fast, intelligent, resilient and collaborative behaviors.Realizing this vision requires research that is informed by large scale experiments involving dozens of interacting heterogeneous vehicles. These experiments require the ability to deploy and coordinate a Wireless Autonomous Systems (WAS) in which an autonomously adaptive groups of agents communicate over a wireless network that is also autonomously adaptive. We therefore propose to develop Penn1, an experimental facility for WAS research in support of several DoD supported research and research-related education activities hosted at the University of Pennsylvania (Penn). The facility will allow for testing, validation, proofing, and development of control, coordination, and planning strategies for heterogeneous teams of wireless intelligent systems. To accomplish this, we are requesting wireless communication infrastructure and a fleet of suitably equipped autonomous ground and aerial vehicles which will be integrated with the GRASP Laboratory existing fleet of aerial, ground, and marine autonomous vehicles into single testbed for wireless autonomous systems (Section 1).Although we are requesting funding for equipment only, the development of Penn1 requires a substantial software development effort that we will fund through leveraged projects. This effort is necessary because access to the experimental infrastructure requires that we commoditize software to simplify the running of experiments. We expect a large user base at Penn that will be interested in leveraging Penn1. Some of them will be interested in testing novel autonomy protocols. Some will be interested in testing novel wireless communication protocols. Some will be interested in both. We will therefore commoditize navigation software (Section 2) and autonomous networking software (Section 3). We will also commoditize an experiment development platform (Section 4).The platform will provide a unique resource for Philadelphia and its neighboring areas. Penn1 will also be beneficial to our many collaborators at the DoD. In particular, the DCIST CRA alliance (W911NF-17-2-0181) will be a primary beneficiary of Penn1 as will ONR Award No. N00014-18-1-2580 and ONR Award No. N00014-19-1-2253 which are joint Naval Research Laboratory (NRL) and Penn projects.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 11, 2020
- Source ID
- N000142012822
Entities
People
- R.vijay Kumar
Organizations
- Office of Naval Research
- United States Navy
- University of Pennsylvania