EquipmentShare reports 37% drop in at-fault DOT recordables after T3 dash cam rollout

EquipmentShare EQPT (U.S.: Nasdaq), a nationwide equipment rental and delivery operator, reports that at-fault DOT recordable incidents per million miles fell 37% year over year after it rolled out T3 dash cameras and the T3 Driver Safety & Performance program, according to a case study published by T3.

The comparison is 2024 against 2025, the first full year after the January 2025 deployment. T3 says at-fault material-loss incidents involving $5,000 or more in damage fell 25% over the same period. T3 describes the rollout as fleet-wide rather than a small pilot.

How the program works

T3 describes a closed loop: detect, validate, coach, and give drivers their own data.

• Event-based video. Cameras do not upload footage continuously. Video uploads when the system detects a qualifying event, such as forward collision, following too closely, phone use or speeding.

• AI flag, human review. T3's AI classifies the event, and EquipmentShare's safety team reviews the footage in context. Events judged false or non-preventable, such as a hard stop caused by another vehicle cutting in, are discarded rather than counted against the driver.

• Coaching on confirmed patterns. When a driver shows a confirmed pattern of risky behavior, the video supports a documented conversation with a manager.

• Driver visibility. Through an app called DriveShield, drivers see their own daily safety score and event video.

Coaching and in-cab alert results

T3 reports that the typical driver saw a 59% reduction in risky driving events after a coaching session. It says 84% of coached drivers improved and 8 in 10 kept that improvement.

For in-cab alerts, T3 compared behavior three months after activation with the pre-alert baseline:

• Phone use and distraction: down 48%

• Harsh driving: down 32%

• Speeding: down 15%

T3 says these reductions have held rather than faded.

What to keep in mind

The source is a vendor case study about a vendor's own customer. It does not give fleet size, miles driven, absolute incident counts or how the baseline was built, and it does not separate the effect of cameras from coaching, alerts or other changes in operations. A rate per million miles can also shift with changes in mileage mix.

Still, the structure is useful as a benchmark. The metric is a standard one, DOT recordable incidents per million miles, and the program pairs human review of AI-flagged events with coaching and driver-facing scores instead of relying on alerts alone.