Comparative Location Testing: An Analysis of Rapid-SOS PSAP Integration and the Effects on Location Accuracy for Smartphone 911 Calls
K. Larsen, R. Ekl, N. Horelik, and R. Mann
Sep 12, 2017|AEDR 2017 Vol. 5 Issue 2|Poster Abstracts, Research Posters
Introduction: Today's wireless 911 location technology is not always precise. As people move away from using landline phones, more and more calls to 911 are being placed from mobile phones, even indoors. In fact, over 70% of 911 calls today are made from mobile phones according, to the FCC. As such, the need exists to improve location accuracy for mobile 911 calls in order to provide fast and reliable 911 response.
Objective: The objective of this study was to evaluate a potential improvement in wireless 911 location technology by performing a pilot test comparing an existing wireless Phase 2 system to a device-based hybrid location system.
Methods: The project team, consisting of NCTCOG, GeoComm, RapidSOS and City of Frisco, TX staff, performed comparative location testing in Frisco, TX in December 2016 with 15 test calls to Frisco 911. Wireless Carrier location from AT&T and Verizon was captured through traditional wireless location mechanisms (Phase 1 and Phase 2 ALI information). In parallel, device-based hybrid handset location was sent to RapidSOS throughout the 911 call via AML mechanisms and captured in the GeoComm 911 mapping product.
Results: Finding 1: Device-based hybrid location highly accurate for more than 90% of calls
14 out of 15 test calls had dispatchable locations obtained from the RapidSOS system. The average distance of device-based hybrid location to ground truth was 16 meters, and the median distance of device-based hybrid location to ground truth was 10 meters. 1 of 15 calls had a location discrepancy to ground truth of more than 50 meters (90 meters).
Finding 2: Wireless Carrier location inconsistent, high discrepancies especially indoors
7 out of 15 test calls had dispatchable location obtained from Wireless Carrier location system.
One of 15 calls did not present Phase 2 information upon re-bid. For the remaining 14 calls, the average distance of Carrier location to ground truth was 202 meters, and the median distance of Carrier location to ground truth was 59 meters. While location accuracy was comparable with the device-based hybrid approach in outdoor environments, in indoor scenarios Phase 2 positioning methods performed very inconsistently.
Finding 3: Wireless Carrier Phase 2 location incurs delay
There is no delay in obtaining device-based hybrid location from RapidSOS after the initial ALI response. Wireless Phase 2 location incurred a delay of average 18 seconds (and median 19 seconds) after the initial ALI response.
Finding 4: Indoor maps add value when Wireless location is accurate
For 9 out of 10 locations where indoor mapping was available, the location was plotted inside the correct room.
Conclusion: This study indicates that location data captured through device-based handset location sent to the RapidSOS NG911 Clearinghouse is more efficient, effective, and accurate than traditional Phase 2 location. In all test scenarios, device-based hybrid location was available by the time traditional Wireless Phase 1 information came in, and the caller location was on average more accurate than traditional Wireless Phase 2 mapping efforts.