Chicago Startup Innovates Gunshot Detection with Smartphone Technology
A startup originating from Chicago is pioneering a novel methodology for gunshot detection that harnesses the capabilities of smartphones as acoustic sensors, with the potential to preserve lives. This initiative follows the city’s termination of its contract with SoundThinking nearly two years prior.
Founded in 2025, SplitSec has engineered an application that utilizes the integrated microphones of Android devices to detect gunfire.
This innovative approach could obviate the necessity for static sensors, akin to those previously deployed by SoundThinking.
“Observing the surroundings, it became evident that during gunfire incidents, bystanders often remain oblivious for several minutes, unsure of what just occurred.
Our aim was to enhance situational awareness promptly,” expressed SplitSec’s founder and CEO, Sri Deivasigamani, in a conversation with Fox News Digital.
“Our focus centered on enhancing personal safety and fostering awareness among individuals, utilizing technology that we all carry daily: smartphones,” Deivasigamani added.
Currently, two complimentary trials are being conducted in Chicago’s 21st and 42nd wards, as reported by Block Club Chicago. There are conversations underway regarding broadening testing to include the 5th, 29th, and 38th wards.
According to Deivasigamani, participants in the trials have provided invaluable feedback. Some users reported heightened instances of false positives due to sensitive microphones, while others experienced significant battery drainage, an issue that SplitSec’s engineers are diligently addressing.
“On July 27 and 28, certain users confirmed the app’s detection of gunfire, affirming they indeed heard shots. This marked a crucial moment for us,” he noted.
Although SplitSec was unable to authenticate the gunfire independently, users validated that the sounds captured by the app were genuine instances of gunfire.
In contrast to the fixed sensors deployed by SoundThinking, SplitSec leverages the microphones and computational capabilities already present in smartphones.
Deivasigamani articulated that advancements in smartphone technology have rendered SplitSec’s innovative approach feasible.
“Such a concept was unattainable a few years back; however, today’s smartphones are AI-enhanced, equipped with substantial computational prowess. We crafted an AI model that operates entirely on the user’s device,” he elaborated.
“Much like your phone responds when you say, ‘Hey, Siri’ or ‘Hey, Google’, our app activates solely in response to gunshot sounds.”
Deivasigamani likened SplitSec to Waze, emphasizing its reliance on both detected sounds and user confirmation, thereby fostering a community-driven network.
Once a gunshot is detected, the app prompts users to verify whether they have indeed heard gunfire. Upon confirmation, SplitSec can disseminate alerts to individuals within a mile radius.
Furthermore, if two or more users concur regarding the same gunshot, the company can triangulate the shooter’s location, thus aiding individuals attempting to evade danger.
“Numerous shootouts have occurred where people inadvertently run toward the source of the gunfire, simply because they lack awareness of its origin,” Deivasigamani noted.
Chicago has a history with gunshot detection technology, having previously employed SoundThinking, formerly known as ShotSpotter, leading to political controversies regarding its cost-effectiveness.
In 2024, the city severed ties with the company, resulting in the Chicago Police Department (CPD) no longer receiving gunshot alerts.
Currently, SplitSec operates independently of CPD, directing alerts primarily to users. However, Deivasigamani indicated that cooperation with CPD or Chicago’s Office of Emergency Management and Communications (OEMC) could facilitate the sharing of an anonymized gunfire map with law enforcement.
The utilization of residents’ smartphones for gunfire detection has spurred concerns regarding privacy, particularly inquiring whether SplitSec engages in continuous monitoring.
Deivasigamani clarified that audio processing occurs entirely on users’ devices, with audio clips being transmitted to the company solely upon user consent after a false positive is reported.
In such instances, users are asked to provide the clip to SplitSec, which utilizes the data to enhance the model’s accuracy.
“We firmly believe that an application of this nature must cultivate user trust while ensuring the protection of privacy,” he asserted.
Source link: Wfmd.com.
