Why 'indoor GPS' for 911 calls is still years away

Although commercial tools for determining the location of mobile phones being used indoor show “significant promise,” they are not yet good enough for first responders to depend on them when answering 911 calls, according to a Federal Communications Commission study.

“While the location positioning platforms tested provided a relatively high level of yield, as well as improved accuracy performance, the results clearly indicate additional development is required,” especially in urban environments, the study’s report concluded.

Controlled tests of three commercial location technologies were conducted by a working group of the FCC’s Communications Security, Reliability and Interoperability Council (CSRIC) late last year in the San Francisco Bay area to establish a baseline of current capabilities. The FCC already has outdoor accuracy requirements for locating emergency cell phone calls to 911, but the technologies being used tend to break down inside buildings.

“As wireless usage increases and as more people are using cell phones indoors (or have abandoned the usage of landline phones altogether) it is becoming clear that the need to accurately locate wireless users in indoor environments is increasing,” the report said.

Accuracy to within 50 meters might be useful in a rural or suburban environment, the FCC said, but in an urban area that could include several buildings. Anything beyond 50 meters is only “general location information.”

The CSRIC working group on 911 location reported last year that there was a lack of objective information available on the performance of location technologies for indoor environments. The tests conducted in November and December helped to fill that gap. They produced a range of accuracy and reliability across the three platforms being tested and across the environments they were tested in. Each technology had its strengths and weaknesses, but in general, urban areas presented the greatest challenges. Because location technologies have focused on horizontal positioning — locating a caller on the ground — getting an accurate vertical location for someone in a multistory building is especially difficult.

“Even the best location technologies tested have not proven the ability to consistently identify the specific building and floor, which represents the required performance to meet public safety's needs,” the report said. “This is not likely to change over the next 12-24 months.”

As the technology improves, implementing it will take time. “Many positioning methods require handset modifications,” the report said. “Integration of these modified handsets into the subscriber base, once the location technology is commercially available, will take years to complete.”

Paying for the technical innovation also could be a challenge. Although public safety is critical, it is not a large enough market to drive commercial development and the wireless location industry is not awash in cash today, said working group co-chair Norman Shaw.

“The market is not one with a lot of revenue potential at this time,” said Shaw, who also is executive director of government affairs at Polaris Wireless, one of vendors in the test. Since the late 1990s “everybody has been saying next year is the year” that location-based services will take off. “But the killer app is not there yet.”

Polaris makes most of its money from foreign governments that use the technology for legal intercept and surveillance, Shaw said. Qualcomm, another vendor in the tests, makes its money from chipsets. The third company, NextNav, uses a dedicated terrestrial network based on satellite repeater towers to provide GPS-like signals, and although the system works well where available it will require significant capital investment to build out its infrastructure.

Still, the tests are an important step in helping the technology mature, and Shaw called it “kind of a miracle” that the vendors, carriers and other stakeholders came together for the venture.

Location testing is logistically challenging and expensive. The Bay Area test cost $240,000, which was provided by participating vendors and carriers. To keep the tests objective, vendors did not take part in selecting the test sites and actual testing was done by TechnoCom, an independent company selected competitively.

Tests were done in four environments: Dense urban, urban, suburban and rural, in San Francisco, San Jose, and adjacent areas. A total of 75 calling sites were chosen in 19 buildings, and at least 100 calls were made from each site on each device being tested. To determine the accuracy of results, surveys established accurate latitude and longitude for each site to within 1 centimeter horizontally and 2 centimeters vertically.

It took eight to 16 hours, in multiple visits, at each of the 19 buildings to get permission to do the tests, do surveys and perform the full test.

Buildings used included the stadium at AT&T Park, home of the San Francisco Giants; the Moscone Convention Center; and the U.S. Court of Appeals Building in San Francisco, as well as other public, commercial and residential buildings.

Shaw said the test results were consistent with expectations, and although the report called for additional development, he said that “with existing technology you can probably replicate the outdoor level of accuracy in an indoor environment.” He said that the FCC could begin consideration of indoor location rules for 911 service based on this level of accuracy this year.

“Although it is not a panacea, it is a step forward,” he said.