What’s next.

Continued progress in reinventing the relationship between people and technology.

Retiring the notepad: The AI in incident reporting

Artificial Intelligence (AI) can ease the process of incident reporting for law enforcement professionals. New voice-recognition tools allow officers to better maintain situational awareness and protect communities more efficiently, without sacrificing accuracy.

Across many industries, we’re reading and hearing about how artificial intelligence will change the future of how we work and automate tasks traditionally performed by humans. From driving cars to predicting sleep cycles, there doesn’t appear to be anything that machines won’t eventually do; and we’re all watching to see what comes next.

With all this buzz, it’s easy to miss the ways in which AI is impacting some important everyday tasks. In Law Enforcement, for example, AI-powered speech recognition technology is helping with one of the most important tasks officers spend hours on each day – completing incident reports.

Now, instead of putting “pen to paper,” officers can dictate an incident report, conduct common lookup tasks – like a license plate lookup – using simple voice commands, or enter data into Records Management Systems (RMS), all by voice. These new voice-recognition tools allow officers to better maintain situational awareness and protect communities more efficiently, without sacrificing accuracy.

So how does AI-powered speech recognition work? Here’s a somewhat simple explanation for a complex technology. Speech recognition technology uses the human voice as the main interface between the user (the person speaking) and the computer (the machine that will automate the task of typing). It’s a highly sophisticated technology that leverages “language modelling” to recognize and differentiate among the millions of human utterances that make up any dialect. Using statistical models, speech recognition programs analyze an incoming stream of sound and interpret those sounds as commands and dictation.

Through machine learning, speech recognition systems learn each word individually as well as which word combinations typically occur. This means that it gets smarter the more it’s used, and it can differentiate between words like “where” and “wear” by understanding their context in a sentence.

Even more important, today’s advances in speech recognition mean it can study the way a specific user talks – learning their accent or the subtle nuances of their voice, and deliver 99% accuracy. And with background noise a major factor in fields like law enforcement, for instance, advances in AI technology mean it also knows to “ignore” ambient noise and focus specifically on the individual speaker.

Advances in AI technology will continue to enhance the work we all need to complete day-to-day. For law enforcement, less time manually creating reports means more time out in the community keeping us all safe. Now that’s some AI worth buzzing about.

Tags: ,

Mark Geremia

About Mark Geremia

Mark Geremia is Vice President and General Manager for Dragon Professional and Consumer and oversees the product and marketing strategy for Nuance's Dragon speech recognition and documentation workflow portfolio. Mark has held various leadership roles within the Dragon business over the last decade, and with his team continues to expand Dragon's reach across enterprise, legal and law enforcement markets, transforming productivity and documentation accuracy for professional individuals and large organizations. Prior to joining Nuance in 2005, Mark held key marketing management positions at both large and small technology companies. He holds a Bachelor of Science Degree in Business Management from Bentley College.