The US National Highway Traffic Safety Administration (NHTSA) has launched yet another federal crash investigation in which Tesla's Autopilot system may have contributed, marking NHTSA's fourteenth probe into potentially Autopilot-related crashes, according to Reuters. Three of those crashes occurred within a three-week span in December, which makes sense given that the ramp-up in deliveries of the Tesla Model 3 has put the Autopilot system in more drivers' hands.

The latest crash to attract a full probe from NHTSA occurred on December 29th in Indiana, when a Tesla Model 3 collided with a parked fire truck alongside the highway, seriously injuring the driver and killing his wife in the passenger's seat.

Another fatal crash that same day, which occurred in Gardena, California, involved a Tesla Model S running a red light and colliding with a Honda Civic at an intersection, killing two of the Honda's occupants. Autopilot was confirmed to be engaged at the time of that incident, while it's unclear yet whether the system was involved in the Indiana crash.

These crashes are helping to fuel an ongoing debate regarding the safety of Tesla's Autopilot system, both with regard to how well Autopilot can detect and avoid stationary objects, and whether Tesla has done enough to ensure that drivers remain alert even while the system is active.

In all, NHTSA has investigated 23 crashes in which some sort of advanced driver-assistance feature is thought to have been a factor, meaning that Tesla's Autopilot system represents an outsize portion of the agency's total special crash investigations.

Like Nissan's ProPILOT Assist, Autopilot requires that the driver keep a hand on the wheel at all times during operation - a safeguard meant to ensure that the driver is attentive while the system is active. But Tesla's system has proven easy to cheat, allowing many drivers to get away with napping or doing other things during operation, and leaving driving entirely up to the Autopilot system.