You are here
LINUS: An Intelligent Digital Assistant for UAM Operators
Phone: (240) 406-5506
Email: bstewart@i-a-i.com
Phone: (301) 294-5220
Email: rbeahm@i-a-i.com
Address:
Type: Nonprofit College or University
Intelligent Automation, Inc. (IAI), along with its teammate University of Maryland propose to develop LINUS, an InteLligent DIgital AssistaNt for Urban Air Mobility (UAM) OperatorS. The key innovation of this project is the development and integration of Automatic Speech Recognition (ASR), Natural Language Understanding (NLU), and Text-to-Speech (TTS) along with their acoustic and UAM developed Language Models to create a digital assistant for UAM Operators. UAM is emerging as a viable alternative to ground and air transportation modes and the UAM Concept of Operations (UAM ConOps) describes a phased-in approach to UAM operations in the National Airspace System (NAS). The roles and responsibilities assigned to operators of UAM will require new technologies and capabilities to achieve this. The IAI Team proposes an intelligent digital assistant (IDA) to assist UAM Operators and Pilots-in-charge in the planning and execution of UAM operations. The core objective of our LINUS project is to develop a natural language artificial intelligence-powered interactive assistant to UAM operators for conducting UAM operations to help manage operator workload. Assistive tasks performed by LINUS will include flight planning, flight following, and contingency management.nbsp; A secondary objective, is to develop a standard API to enable integration with future backend AI-based intelligent information systems. For this proposed project, we intend to leverage prior work that developed a chatbot for comfortably discussing and reporting symptoms, and Intelligent Digital Assistant (IDA) project that developed technology to ease operational pressures on Sailors interacting with complex Navy Combat Systems. IDA integrates state-of-the-art ASR, NLU, and TTS architectures coupled with domain specific neural acoustic models for ASR and domain specific language models for ASR, NLU, and TTS.
* Information listed above is at the time of submission. *