Only Samsara provides the full picture of risk: Announcing Samsara’s latest AI detections

Reducing accidents and improving driver safety requires visibility, precision, and proactive intervention. That’s why Samsara continues to expand its suite of AI-powered detections, giving fleets the most comprehensive view of driver risk in the industry.

Samsara AI detections, analyzed within the Samsara Dashboard, provide an unmatched level of risk assessment. First, our AI detects more than 30 risky driving behaviors, blind spot hazards, and external risk factors like weather, road, and traffic conditions. But there’s more. Because our AI is trained on the world’s largest Connected Operations dataset—20 trillion data points, 90+ billion annual miles, and 180+ billion minutes of real-world footage—it can differentiate between one-off instances and patterns of behaviors, which incidents occurred during dangerous or high-risk conditions, and which drivers are most at risk. 

These latest detections provide safety teams with new precision in identifying, coaching, and preventing risks—helping them act before incidents occur.

Protecting pedestrians, drivers, and communities

Pedestrian collisions are on the rise—pedestrian deaths increased 12% between 2019 and 2023. With a single fatal crash involving a medium or heavy truck estimated to cost $7.2 million, proactive detection is essential to reducing exposure and liability.

Pedestrian Collision Warning extends AI visibility to the vehicle’s surroundings, detecting pedestrians and other vulnerable road users such as cyclists, scooter riders, and road workers who may be at risk.

If such a road user is detected nearby with potential for impact, the system issues an in-cab audio alert to warn the driver, giving the driver a chance to correct course before a collision occurs.

Samsara’s AI-powered Pedestrian Collision Warning extends all around the vehicle. Our AI Dash Cams monitor the road ahead, while AI Multicam extends detection to the sides and rear—helping prevent incidents in real-time.

Coach USA, one of the largest transportation companies in North America, has reduced incidents through focused safety initiatives and has recently deployed this feature. “Through focused safety initiatives and our partnership with Samsara, we’ve achieved a 92% reduction in preventable incidents,” said Jason Louis, VP of Safety. “Samsara’s AI Multicam and Pedestrian Collision Warning will enhance our efforts to promote safer roads, move toward zero incidents, and remain a trusted partner for communities nationwide.”

Supporting drivers parked roadside

Parked vehicles on high-speed roads pose a significant safety risk. With limited reaction time and narrow shoulders, drivers are exposed to high-impact collisions that can cause severe injury, loss of life, and significant liability. Litigation can cost millions, with one recent case settled for $3.5 million. Samsara’s Roadside Parking Detection helps fleets identify these high-risk situations and intervene before they escalate.

If a vehicle is parked on the shoulder of a high-speed road for an extended period—typically more than 10 minutes—the system automatically flags the event. Video footage and context are sent to the Safety Inbox, enabling managers to quickly assess the situation and act.

With this visibility, safety teams can check in with drivers to confirm their well-being, determine the reason for the stop, and coach them on proper procedures such as deploying warning devices. Whether the cause is a mechanical issue, medical concern, or non-compliant rest break, timely outreach helps prevent roadside collisions and keeps drivers safe and supported on the job.

Validating speeding data with AI analysis of posted signage

Accurate detections are essential to fair coaching and driver confidence. Samsara’s Sign-Verified Speeding uses computer vision to validate speeding events, ensuring alerts are grounded in fact.

When a potential speeding event is flagged, Samsara validates the event against our verified speed-sign database—a library built by state-of-the-art computer vision models that capture speed-sign images across our extensive dashcam network. If a verified sign exists for that specific road segment, the matching sign image and a link to the Google Maps Street View appear alongside the event, providing clear, credible evidence for coaches and drivers alike.

Samsara’s Sign-Verified Speeding brings unmatched precision to speeding detections. This saves managers time, eliminates the need for manual verification, and builds trust that every event flagged is accurate and actionable.

Upholding company standards and ensuring compliance

Consistent, automated visibility helps fleets maintain safety standards and reduce compliance risk. Samsara’s latest AI detections make it easier to monitor critical behaviors—both on the road and in the cab.

Railroad Crossing Detection identifies when a vehicle fails to stop before crossing railroad tracks, regardless of the presence of a train, which is often required by law or company policy, helping fleets prevent highly fatal and preventable accidents. Driver error is the probable cause in 94% of vehicle crashes at highway grade crossings, and this detection provides automated oversight where it matters most. For regulated fleets like school buses and hazmat carriers, it also helps simplify compliance with DOT and industry-specific requirements to stop at every crossing.

Passenger Detection adds another layer of accountability. Using inward-facing camera footage, Samsara identifies when an additional person is present in the cab during a trip and surfaces the event in the Safety Inbox. This helps managers enforce company policy, prevent unauthorized use of vehicles, and reduce liability associated with unapproved passengers.

Together, these detections give fleets the visibility to uphold company standards, automate compliance, and protect drivers and assets from unnecessary risk.

All of these new AI detections are available now and can be activated directly in your Samsara Dashboard—helping you take another step toward zero incidents and a safer, smarter operation.