Driver judgment, lack of awareness, and fatigue remain the leading causes of large truck crashes, according to U.S. Federal Motor Carrier Safety Administration (FMCSA) data shared during a Samsara safety panel.
The session, part of Samsara’s Beyond 2025 conference in San Diego, Calif., outlined how AI-powered safety tools are being deployed to help fleets identify and mitigate these risks in real time.
FMCSA data show that 62% of fatal large truck crashes involve driver judgment — including speeding and close following. Another 34% relate to limited driver awareness, and 13% are tied to impairment, primarily from fatigue or physical issues.
Speeding and close following
Speeding is involved in 30% of crashes, and 28% of all motor vehicle crashes are rear-end collisions. This is why Samsara has introduced evidence-based speeding detection by equipping its dashcams with speed sign recognition. The feature captures posted limits in real time and pairs them with speeding alerts, allowing safety managers to review image-based proof and reduce ambiguity in coaching conversations.
“In the near future, if a speeding [event] occurs in close proximity to the speed sign, what you’ll be able to do is view images of that speed sign on the event. You’ll be able to see exactly when those images were taken, not just from vehicles historically, but across all vehicles in Samsara, anonymized,” explained Lucas Van Houten, director of product management at Samsara. “If you click on this image, it will take you directly to Google Street View, where you can then do the sign for yourself…You’ll no longer have to have arguments about, ‘Oh, it wasn’t a sign. I didn’t see the sign.’ It’s pretty clear evidence in these cases.”
As an extension of the newly announced StreetSense feature, new weather-adaptive safety threshold capabilities allow to tighten speeding and following distance alerts when the system detects rain, fog, or other severe conditions. AI uses anonymized dashcam footage from nearby vehicles in the Samsara network to determine whether a vehicle has entered a verified weather risk zone.
“We optimize the whole weather alert area. We look at the images coming back from the camera of each vehicle and the cameras of other vehicles on the network in very close proximity to your vehicles, so we can know with high accuracy at the few hundreds of meters level exactly where it is raining and where the severe weather is occurring,” Van Houten said.
Mobile use and blind spots raise awareness issues
Driver awareness is another major concern. According to Samsara’s own research, 79% of drivers said they experienced a “close call” due to distraction in the past year. Mobile phone use increases crash risk by 23 times, and blind spot-related errors contribute to 14% of large truck crashes.
To address this, Samsara says it has improved its AI model for mobile phone detection, claiming it is now 99% precise. Multicam systems with AI detection — announced during the Beyond conference as well — are also being rolled out to give drivers side and rear visibility, especially in urban settings or high-risk zones with cyclists and pedestrians.
Fatigue and driver distraction contribute to collision risks, too. According to Samsara and AAA Foundation research, drivers who miss just one to two hours of sleep can double their crash risk. In a 2024 survey by Nationwide, 61% of commercial drivers reported longer working hours. Samsara’s internal analysis found that when drivers split their shift into more trips or take breaks, crash risk falls by about 9%.
Samsara’s fatigue model uses 17 behavioral indicators — including yawns, blinking, and head movement — to generate a fatigue score that evolves over the course of a trip. A “See All Signs” feature now compiles video clips from the previous 30 minutes to give supervisors more context when deciding whether to intervene. The system can alert managers via the dashboard, SMS, or email if fatigue indicators cross a threshold.
However, just installing new tech without intention won’t help fleets. Michelle Dixon, senior director of transportation safety and compliance at Sysco, said that safety systems must be integrated into a fleet’s broader culture, not just deployed as oversight tools.
“I would argue that the goal is not to be proactive — the expectation is to be proactive, and the goal is to be predictive,” she said.
