A Voice Interface for Mobile Data

Posted On Feb 23, 2016 3:59:44 PM by Advanced Public Safety

One of the unfortunate downsides of the mobile data era is an increase in the number of wrecked police cars. In a perfect world, with perfect officers, everyone would pull to the side of the road, put the car in Park, and set the parking brake before typing on their mobile data terminal’s keyboard or attempting to read any of the responses that appeared on the display. But, life is not like that, and nearly every agency with mobile data capabilities has experienced an incident where an officer has driven into a tree or another vehicle while their attention was directed at the computer in the car.

A Florida company called Advanced Public Safety has introduced a product that may reduce some of these collisions. Virtual Partner Voice Response™ picks out the critical data from replies to law enforcement database queries and converts them to spoken words, audible without having to look or otherwise interact with the computer. Especially urgent messages, such as those advising that a vehicle is shown to be stolen, are prefaced by an alert tone and a read-back confirmation of the license plate number.

Generic text-to-speech applications have been around for years, but they aren’t very practical for mobile data. In a text-to-speech application, a synthesized voice reads every word on the computer display with a tone and cadence that is often difficult to understand. Just the fact that every word is read back can be frustrating. For instance, when we read a Web page, we more or less ignore the headings and other information that appears at the top of every page, and skip down to the text that we’re interested in. In a text-to-speech environment, the user gets every word read to him, and it might take a while to get to the part that he wanted to hear.

Virtual Partner Voice Response doesn’t use a synthesized voice. Instead, every word is taken from a library of recorded sound files that resides on the host computer. The software matches the appropriate sound file with the words that appear on the display, and plays them in that order. The result is a voice that is much more natural sounding and easy to understand.

The real innovation, however, is that the voice ignores the extraneous information that accompanies most replies to database inquiries. The agency ORI text, the time and date, and other details are skipped over, and the voice reads back only the stolen/clear status of the license plate, the make and model of the vehicle to which it is registered, and the name, license status and wanted status of the registered owner. If this information is other than routine— if the vehicle is stolen, the license is suspended, or the driver is wanted— an alert tone will sound and the original inquiry will be read back so that the operator can confirm that the right information was sent – all without taking his eyes off of the violator.

Advanced Public Safety also produces software that expedites the completion of traffic citations and accident reports by extracting the data contained on the driver’s license magnetic strip and using it to fill in the blanks on the citation form. The appropriate violation is chosen from a drop-down list that limits errors stemming from mismatched statute numbers and captions.

Virtual Partner Voice Response is designed to function with mobile data software from VisionAIR, Aether, SCA Motorola, OSSi, New World Systems, HTE and D.M. Data, as well as others. Demos for each of these software packages are available on their Web site.

Topics News Releases