I-Corps: Autonomous Unmanned Vehicle Ecosystem Technology

Cuny Queensborough Community College, Bayside NY

Investigators

Abstract

The broader impact/commercial potential of this I-Corps project is the use of autonomous robotic systems to save lives across multiple disaster levels by multiplying human efforts in areas such as disaster response, extending to reduced response times and improved rescue efforts during disaster relief. Using autonomous systems, such as drones, that run software which allows for the use of biosensors and on-board machine learning, natural disaster responders can find survivors faster and reduce casualties. The technology understands mission objectives (intent-driven), coordinates with others (collaborative autonomy) and is voice enabled (natural language conversational interface), allowing it to work well with both humans and machines. These and other features lend this technology to broad commercial applications, offering the best strategy to mitigate multiple disaster levels - at the district or provincial level by assessing strengths and vulnerabilities to various hazards; at the community level like search and rescue during a flood or fire; and at the household level like ambulatory drone service care for the aging. This I-Corps project will be focused on translating scientific research to commercialization through lessons on the business model canvas and extensive customer discovery. Our research has involved hands-on technical experimentation using deep/machine learning software for prototype development and literature reviews on a variety of software capabilities. Using a camera with machine learning, we were able to identify a human with most of the body obscured. Hardware acts as transport for the optical and acoustical biosensor package which can take action autonomously by utilizing artificial intelligence (AI) technologies extending to: realtime onboard object detection capable of identifying a person in rubble, even when the person is partially obscured (Machine/Deep Learning/edge), offline cloud-enriched datasets for cloud-sourced analysis, voice recognition/conversational natural language processing/speech synthesis. Blockchain enabled autonomous swarm management technology on multiple robotic devices will enable multiple machine collaboration. All sensor data is time-stamped and stored in the cloud where the AI technologies are used to enrich the data. Enriched data is made available offline via the internet to thousands of volunteers for human review. Results from our research show that the technological capability exists, the software and hardware components combined meet a need established through our initial customer discovery. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.