"When in Doubt, Find Out" – The Call

January 3rd, 2017

Posted In: Industry Insights

When in doubt, find out

Originally published in The Call Magazine, Fall 2016

Wireless integrity testing helps a South Carolina county discover the root cause of its 9-1-1 call-routing problems

By Dave Sehnert and Pat Thompson

Shortly after 4 a.m. on the morning of December 29, 2014, Shanell Anderson, a newspaper delivery supervisor in the Atlanta metropolitan area, was covering for a delivery driver who had called in sick. Anderson was driving an unfamiliar route, and as a result took a wrong turn and inadvertently drove her car into a pond.

The car began to sink, but Anderson kept her composure and placed a 9-1-1 call. According to media reports, Anderson’s call hit a cellular tower in Fulton County, so it was fielded by a public safety answering point (PSAP) in Alpharetta, Ga.; unfortunately, Anderson’s emergency was playing out just a few yards into adjacent Cherokee County. Further complicating matters, Alpharetta’s computer-aided dispatch (CAD) system had no record of the intersection provided by Anderson, even though it was within walking distance of the county line. This is not unusual, as PSAPs often do not have Geographic Information System (GIS) data for adjacent counties loaded into their CAD systems.

It took roughly 20 minutes to unravel the mystery and transfer the call to a Cherokee County PSAP, but by then it was too late for Anderson. Though she was still alive when firefighters pulled her unresponsive body out of the pond, she died a few days later in a local hospital.

Location, Location, Location

The tragic circumstances that led to Anderson’s death underscore the importance of wireless location accuracy, particularly for 9-1-1 calls.

“Location is our biggest priority,” said Jim Lake, director of the Charleston County Consolidated 9-1-1 Center, which dispatches for 21 law enforcement, fire and emergency medical services (EMS) agencies in the county—including the Charleston County Sheriff’s Office and the city of Charleston’s police and fire departments—and which handled more than 341,000 9-1-1 calls last year. “If we don’t know where the caller is calling from, we can’t send emergency response.”

When 9-1-1 is called using a wireline device, a precise location can be determined because the pre-validated address for the handset is maintained in a regional database. Things are not as precise in the world of wireless 9-1-1 because of device mobility.

Based on parameters established by the Federal Communications Commission (FCC), wireless carriers must provide to PSAPs the telephone number of the device that placed the emergency call and the location of the cellular site or base station that transmitted the call, which is known as Phase I location. In contrast, Phase II location requires the wireless carrier to provide the latitude and longitude of the calling device at the time the call was placed, within a range between 50 meters and 300 meters, depending on the technology used. Handset-based technology, i.e., Global Positioning Satellite (GPS), must provide location accuracy in a range between 50 meters and 150 meters; meanwhile, network-based technology, i.e., trilateration, must provide location accuracy that ranges between 150 and 300 meters. With the 2015 update, the FCC rules also mandate the amount of time that the carriers have to provide such information to the PSAP when a wireless call is placed, or a time to first fix (TTFF) of 30 seconds for outdoor calls. The wireless carriers will have new wireless location accuracy thresholds to meet in April 2017, with incremental increases to the level of accuracy over the following five years.

The Reality of Emergency Calls

The location standards noted above have existed for quite some time. In fact, many PSAPs and wireless carriers have addressed this location issue in the past. Yet, with the ever-increasing number of wireless emergency calls, new cellular sites being added to the network, and changing demographics and jurisdictional boundaries, the previously resolved location issues are reappearing. Said another way, a problem once thought to be significantly resolved is back, and the risks associated with it are unacceptable given the public’s reliance upon wireless communications.

Most wireless 9-1-1 calls today arrive at the PSAP with Phase I location information. This is not particularly helpful in pinpointing the caller’s location, as the caller may be anywhere within a three-mile radius (or more) of the cellular site that is handling the call. Consequently, telecommunicators are trained to ask the question, “Where is your emergency,” in addition to other questions that may help them identify landmarks or geographical markers.

When callers can’t provide their location—for example, tourists who are unfamiliar with the area that they are visiting—the telecommunicator who fielded the call can make a series of Automatic Location Information (ALI) database queries known as rebids to obtain Phase II location information. While rebids are effective—and necessary when the caller is on the move—each requires 30 seconds to execute and as many as three might be needed. This is precious time when lives are on the line. Ideally, every wireless 9-1-1 call would provide accurate dispatchable location information from the outset.

There are many other things that can go wrong with a wireless 9-1-1 call. Chief among them are misroutes, i.e., when a 9-1-1 call is sent to a PSAP that is not responsible for the emergency response. This typically occurs along jurisdictional borders, and for several reasons. For instance, the cellular antenna is divided into several cellular sectors, with each sector assigned to a specific PSAP. However, call routing often is made more complicated by the fact that these sectors often are many miles wide and deep; as a result, they can cover multiple PSAP jurisdictions, especially in urban areas or near borders.

In addition, antennas sometimes are misaligned, routing databases become out of date or contain errors, or the cellular sector assignments have been incorrectly provisioned by the wireless carrier—or they need to be updated because circumstances have changed—all of which can cause a call to be misrouted. When such events occur, the receiving PSAP should be able to transfer the call to the correct PSAP; however, such transfers eat up precious time, delaying emergency response. And in rare instances they don’t happen in time, as was the case for Shanell Anderson.

Not long ago, Lake sensed that things were amiss in the Charleston County Consolidated 9-1-1 Center. The big red flag was that about 13 percent of wireless 9-1-1 calls placed in the county were misrouted. That’s a high number—under normal circumstances, only two to three percent of such calls should have been misrouted.

There was other anecdotal evidence. For example, personnel test the Center’s various 9-1-1-related software using their own wireless handsets every shift; depending on the time of day or day of the week, they were getting all sorts of different locations for where they were calling. Also, some of the locations being reported were close to what the database was indicating for Phase II locations, but given the personnel’s knowledge of the area, they didn’t think it was close enough.

However, when working with the wireless carriers, merely thinking that a problem exists isn’t enough. Indeed, the carriers tend to be dismissive about alleged problems unless confronted with hard facts. But there were other important reasons for launching a fact-finding mission, according to Lake.
“The wireless carriers, National Emergency Number Association and the FCC are moving toward changes in the rules and technologies concerning wireless location,” Lake says. “We needed to have a baseline so that we could discern whether those changes, when they finally occur, actually improve our location accuracy.”

What They Learned

The county hired the South Carolina Research Authority (SCRA), which in turn hired Mission Critical Partners to conduct wireless integrity testing. While 50 locations were identified across the county, tourist areas were targeted, because callers in such areas typically will not know where they are, and that makes 9-1-1 calls placed by them more difficult to process. Charleston County has a great many tourist attractions, including state parks, forested areas, game lands, beaches, historic districts, and plantation areas, as well as many coastal islands. Three waterway areas along the coast, where wireless 9-1-1 call activity tends to be higher, were targeted in particular.

Four cellular phones, one from each of the major carriers, were used during testing, and a statistically valid testing model that is replicable was applied to determine accuracy. The testing took two weeks to complete and revealed several interesting facts. First, it was discovered that several dozen tower addresses were inconsistent, mostly in situations where multiple carriers are collocated at the same tower site; in some cases, different spellings and/or sector directionals were used, or NENA standards were used in some cases and U.S. Postal Service standards in others. Such discrepancies need to be addressed by the PSAP and wireless carriers.

Second, when some 9-1-1 calls roamed onto another carrier’s network, the ALI bid response identified the handset as a non-service-initialized (NSI) device, which means there was no callback number or any other location information provided for the calls. The FCC mandates that NSI devices, i.e., those that no longer are associated with a wireless service plan, still must be able to place a 9-1-1 call. While it is not unusual for a call to roam onto another network when coverage issues exist, the fact that these calls were identified as coming from NSI devices when they clearly were not is an anomaly; Mission Critical Partners is working with the affected wireless carriers to gain insight and a technical understanding as to why such results would occur.

Third, the testing indicated that some unidentified cellular sites may exist in the county, and this will require further research.

The most critical finding, however, was that the high percentage of misroutes was due to the fact that many cellular sector assignments need to be re-provisioned via a wireless carrier/PSAP collaboration, because the population, as well as available emergency resources, has changed dramatically since the antennas first were provisioned.

“The question is, who actually should be receiving those calls now that the population has shifted,” Lake said.

Accurate routing of 9-1-1 calls is the lynchpin of effective emergency response, and PSAPs need to work collaboratively with the wireless carriers to find a solution when anomalies exist. The ability to do so depends on empirical data—which the wireless integrity testing conducted by Charleston County provided.

Dave Sehnert and Pat Thompson are technology consultants for Mission Critical Partners, Inc., a public safety communications consulting firm headquartered in Port Matilda, Pennsylvania. They can be emailed at davesehnert@missioncritical.3twenty9.net and patthompson@missioncritical.3twenty9.net, respectively.