Public safety departments today face very real resource challenges. The emergency call volume of most departments has risen due to population increases, annexation of adjacent communities, migration patterns, aging populations with increased medical needs, department mergers and so forth. In many of these cases, however, budgets haven't increased proportionately, and dispatchers and first responders alike are being tasked with handling a larger call volume without staff increases. This increased stress can lead to higher employee turnover at 911 communications centers and stress-related illnesses for firefighters and EMS workers.

In addition, as the same number of units are required to respond to a greater number of calls, response times become every more critical. Automated emergency dispatching is a practical and effective means to address and mitigate these challenges.

Until fairly recently, the “talking computer” technologies behind automated dispatching used in public safety applications left much to be desired in the way of clarity, natural vocal sound and flexibility. But developments over the past 10 years have yielded advances in computer memory and processing power, which in turn have led to significant improvements in artificial speech technology — a memory- and processing-intensive computer application. The improved speech technology represents an available, affordable and practical way to enhance the speed and effectiveness of public safety dispatching. Coupled with controls that can automatically handle lighting, alarms and more in fire stations, such systems offer significant improved functionality and efficiency over existing computer-aided dispatch systems.

Less stress, more productivity

Automated dispatching is a separate, computer-based system that works hand-in-hand with a public safety department's CAD system. It separates and automates the act of speaking the dispatch and automatically generates and routes appropriate text and audio messages to the correct fire stations or rescue crews in the field. It does not take away a dispatcher's job; rather, it makes a dispatcher's job less stressful and increases productivity.

Although automated dispatching system functionality will vary from manufacturer to manufacturer, such systems follow similar routines. First, the public safety department's CAD system generates an emergency message based on the incoming emergency call, including identification of the address corresponding to the emergency. The dispatcher reviews and confirms the message and the rescue units selected for response. Once these aspects have been confirmed, the dispatcher commits the unit or units with a single keystroke or mouse click.

Following the confirmation, the automated dispatching system automatically sends the dispatch information electronically to the station or units designated for response. When more than one dispatch data path is available within the public safety department's alerting infrastructure, the more sophisticated automated systems will select the fastest data path available at that moment. Once the dispatch arrives at a station, it's received by a computer that converts the dispatch from alphanumeric text to an audio dispatch that is broadcast in the station and also can be broadcast via the department's radio system. Dispatches to field units are similarly received via mobile computers in vehicles.

The more sophisticated systems can send multiple dispatches to multiple locations in milliseconds, which is of particular benefit when incidents involve large fires or accidents that require multiple units from multiple stations to respond. Simultaneous dispatching is key to generating faster response times.

Automated dispatching facilitates faster responses and reduces stress to both dispatch staff and responders. With automated dispatching, there's less delay in sending the dispatch than if it's done manually.

In addition, the automated dispatching system features not only a natural human-based digitized voice, but also one that is always clear and accent-neutral. The consistency and clarity of the voice offered by advanced, automated dispatching systems results in fewer “please repeat” requests. There also is less likelihood of errors resulting from a locations being misunderstood; if a dispatcher has a strong regional accent, it is entirely possible that a responder might misconstrue “Sierra Street” to be “Sahara Street,” with disastrous consequences if the two streets are across town from each other.

Today's voice technologies

Automated dispatching is based on text-to-speech processing, which is the automated synthesis of speech from text. In essence, a computer “reads” text and then speaks it out loud. Within the realm of text-to-speech processing there are two distinct approaches: synthesized speech and concatenated speech. Each approach has specific pros and cons.

Synthesized speech is software programming based on predetermined phonetic and language-based algorithms. The computer reads the text and applies the rules as it generates the audio output. A synthesized speech system generates individual speech sounds electronically, mimicking the human voice.

Positive aspects of this approach to text-to-speech include:

• No audio database is needed

  This type of system truly reads the text. There is no need for a library of pregenerated sounds.

• System can be less expensive

  A synthesized speech system can cost less than a concatenated system and takes less time to implement.

• Less demand on system memory

  Synthesized speech systems require relatively little disk space, less RAM and less processing power because they don't use a prerecorded audio database.

The drawbacks of the synthesized speech approach — particularly for a mission-critical application like public safety dispatching — can be significant. Negative aspects of the technology include:

• Robotic sound

  Synthesized speech can sound unnatural and robotic. It often lacks cadence (natural, conversational rises and falls in pitch), pacing and emotional inflection.

• Does not work well with special pronunciations

  Because this type of system reads the text and applies the same rules each time, it can't take into account rules of different regions. For example, the street name “Houston” is pronounced “hew-ston” in the South, and “how-ston” in the East. And synthesized speech software can run into problems attempting to handle a more complex word or spelling, such as “Kruschchev” because of the “schch” portion of the word.

• May mispronounce words when following phonetic rules

  A synthesized speech system pronounces words phonetically, and it's quite common for a synthesized speech system to pronounce the word “Chihuahua” as “Chi-hoo-ah-hoo-ah.” Some synthesized speech systems do allow for manual programming to better pronounce problem words, but typically each problem word must be loaded into the system manually.

• Can't distinguish homographs

  Homographs are words that are spelled the same but have different meanings and pronunciations. Consider the phrase, “I read the newspaper every day.” Should the word “read” be pronounced in the present tense as “reed” or in the past tense as “red”? Synthesized speech would have trouble distinguishing between the two.

• Limited ability to convey emphasis or emotion

  Part of what makes the spoken word more understandable is the use of emphasis and emotion to convey the intent of the communication. Consider the many different meanings the phase “Don't do it that way” might take on, depending on which work is emphasized. “Don't do it that way” could get an entirely different response from “Don't do it that way.” Synthesized speech can convey only limited emphasis and emotion.

Concatenated speech is software programming that makes use of a prerecorded audio database. With this approach, a trained professional speaker records the necessary vocalizations. In public safety applications, that would include numbers, street names, prominent landmarks, unit numbers and incident types. This information is stored within an automated dispatch system computer as a database of audio sounds. When the system reads the text, it then reassembles the correct words and audio phrases to form spoken sentences.

There are several positives associated with concatenated speech, including:

• Clear and easy to understand

  Because concatenated speech is based on a pre-recorded human voice created in a noise-free studio environment, the voice quality is exceptionally clear.

• Sounds natural

  Because its audio components are derived from a real human voice, concatenated speech sounds real — so real that it can be difficult to distinguish between a concatenated computer voice and an actual person speaking. In addition, concatenated speech technology can re-assemble words and phrases with natural rhythm and pitch, creating a conversational rather than robotic quality to the voice.

• Little chance for mispronunciation

  Concatenated speech offers little chance for pronunciation errors because it's based on a prerecorded audio database.

• Cost-effective

  In the past, this technology option would have been more expensive than synthesized speech because it does require more memory and processing power to store and manipulate the audio database. Today's computer systems, however, can easily accommodate the storage of an audio database.

The drawbacks of concatenated speech technology include the fact that an audio database required. This technology requires that a trained professional speaker pre-record all the words and phrases the system will need. In addition, Concatenated speech requires significant computer memory space and processing power. This technology does require storage to support the audio database. For example, a 15,000-entry audio database would require roughly two gigabytes of storage, which might pose a problem on an older computer, as might a slower chip set. Even entry-level PCS today, however, are being built with 40 GB of memory and fast central processing units.

Keys to automated dispatch

When considering the potentials of an automated dispatching system for your department, keep the following points in mind:

Simultaneous dispatching capability should be a priority

This function is one of the top reasons why automated dispatching reduces stress and response times. Simultaneous dispatch capability should be at the top of your list of specifications when you are considering an automated dispatching system.

Go concatenated

Unless or until significant improvements in the accuracy and clarity of synthesized speech become available, concatenated speech technology is optimal for public safety applications.

Centralize high-level system functions at the 911 call center

Distributed computing makes sense for low-level functions, but it does not when high-intelligence functions are involved. Providing high-level capability to every station will increase the overall cost of the system. Automated dispatching systems are most cost-effective when their high-level intelligence functions are centralized at the 911 communications center.

Seek a software-based system

Most automated dispatching systems are driven by software and run on hardware. To ensure cost-effective, long-term system maintenance and facilitate adding more stations, the audio portion of the automated dispatching system should be software-based rather than stored on a chip. This also can be important with respect to upgrading. In addition, a software-based system allows for remote updating. A hardware-centric system stored on chips requires sending a technician to each station to manually remove and replace chips. Obviously, the more stations covered, the more time and potential expense may be involved in an upgrade.

Automated dispatching means less staff stress, not less staff

When the subject of automated dispatching comes up, some dispatchers hear the word “automated” and think it means losing a job to a computer. The reality is that automated dispatching systems simply make their lives easier once the systems are installed. Don't allow bad information to sidetrack an initiative. Consider involving dispatch staff in the process from the beginning.

---

Courtney DeWinter is a freelance writer.

Sidebar