What Behaviors Cause Crashes?

Over 8.5 million people have CMT’s crash technology on their phones today. The technology spans over 20 programs across the world, implemented with different user experiences.

Many of the programs are with insurers who use the real-time crash detection technology to help customers who opt-in by sending emergency services to the scene of the crash. Some insurers are still experimenting with the user experience after a crash. Today, they’re focused on using the data to accelerate the claims process for customers and to make it more efficient for their own operations. Many companies beyond insurance are using CMT’s real-time crash detection to help drivers, such as automakers, wireless providers, and home security companies.

Drivers write us frequently to tell us how our technology has improved their lives. Recently, a man wrote us about his experience with emergency crash assistance. He was driving on a road in the country and had a hard attack. He fell unconscious and crashed his van into a tree at 17 mph. CMT’s technology detected the crash and alerted emergency services. They dispatched paramedics to the crash location, who found the driver face down in a ditch with no pulse and no heartbeat. Paramedics revived him and sent him to a nearby medical center by helicopter. In his note to us, he said: “Your app literally saved my life.“

CMT’s crash detection technology leverages advanced signal processing and machine learning techniques to identify crash events. This proprietary AI-powered approach allows CMT to detect crashes faster with higher accuracy and at lower impact speeds. When paired with CMT’s IoT Tag device, which drivers stick to their windshield, this technology detects crashes at ultra-low speeds, including when the car is parked.

With over a decade of experience in risk analysis, paired with crash analysis and AI-driven reconstruction, CMT has the ability to build the largest naturalistic driving assessment, enabling CMT to have a highly scaled view of the cause of crashes. The scale provides CMT with a deeper understanding of how individual risk factors like distracted driving impact the likelihood of crashing.

One of these findings shows just how prevalent crashes caused by distracted driving are. In a study across multiple US auto insurers, we analyzed how often drivers were using their phones the minute before they crashed. We found that 34% of drivers who crashed had used their phones in this time frame.

Crash insights from over 1.8 million drivers in 2023 confirm what we reviewed in the section before on crash frequency and severity. The difference here is that we aren’t dividing drivers by risk segments or calculating frequency or severity lift. Instead, we segmented 1.8 million drivers into two groups: those who crashed in 2023 and those who didn’t. We then analyzed how often they engaged in risky behaviors. For example, drivers who didn’t crash in 2023 interacted with their phone while driving 2.8% of the time. Among drivers who crashed, this figure was close to 4.6%, a 62% increase.

Among the types of distraction events, hands-free calls had the lowest impact in increasing the likelihood of crashing. However, drivers who spent more time making hands-free calls had a 52% higher chance of crashing. Handheld phone calls had the biggest impact on crash risk among this group of drivers, increasing the chance of crashing by 100%.

We also see that drivers who more frequently use hands-free phone features have a higher chance of crashing. This metric indicates that the screen is activated or that audio is playing, and that the driver isn’t handling their phone. It covers a broad number of use cases, such as using navigation apps or listening to music, podcasts, or audiobooks. What this particular metric reveals is that increases in any distraction can lead to a higher chance of crashing, similar to the National Institutes of Health’s finding that eating can increase the chance of crashing by 3X for younger drivers. Download 2024 The State of US Road Risk Report