DHS Develops Hands-Free Communication Technology for First Responders

The Department of Homeland Security Science and Technology Directorate has unveiled a new hands-free communications technology for first responders called Direct Artificial Intelligence System Interface.

DAISI is powered by automated speech recognition technology, which the DHS S&T developed with Johns Hopkins University Applied Physics Laboratory and ASR specialist Think-A-Move.

The tech outperforms current speech recognition systems owing to voice-activated capabilities that enable first responders to communicate effectively even in noisy operational environments, DHS S&T said.

APL Project Manager Julee Rendon noted that DAISI’s voice recognition capability relies on machine learning.

“It’s really about the language processing, the acoustic modeling and the noise filtering,” she said.

With DAISI, users can perform actions to navigate roads without lifting a finger. The communications tech can answer address queries, provide alternate routes and pan and zoom throughout the map through voice instructions alone.

In fire incidents, DAISI can field hydrant location queries and generate friction loss calculations for fire hoses.

The technology is also capable of accessing the Wireless Information System for Emergency Responders, a system designed to assist emergency responders in hazardous material incidents. WISER offers information on hazardous substances like physical characteristics and containment and suppression advice.

Additionally, DAISI has features for on-scene report logging and transcription, ensuring that first responders are able to make comprehensive and thorough after-fire critique and fire investigation reports.

The concept for DAISI dates back to 2018. It was developed as part of an April 2018 request for proposals and later subjected to a three-year prototype test with Maryland’s Howard County Department of Fire & Rescue Services.