Ford Mustang Mach-E using BlueCruise at time of crash: NTSB

Crash scenario and BlueCruise context

• Crash involved a Mustang Mach‑E with BlueCruise active, striking a stopped CR‑V at night in the center lane of a high‑speed Texas freeway.
• Witness reports: CR‑V had no tail or hazard lights; a preceding car saw it and swerved, then saw the Mach‑E hit it in the mirror.
• Several commenters note this “peek‑a‑boo” setup (obstacle occluded, then revealed too late) is hard for both humans and automation, especially at 70–80+ mph.

Human vs system responsibility

• Some argue a human “easily” could have made the same mistake; others counter that a preceding human did avoid it, while the system and its supervising driver did not.
• Repeated emphasis that BlueCruise is Level 2: the driver is still legally and practically responsible, and manuals warn about stopped objects.

Limits of current ADAS / self‑driving

• Many say today’s systems are like supervising a novice driver: less workload reduction than advertised, and you must stay fully alert.
• Radar-based ACC/ADAS often ignore stationary objects at highway speed due to false positives from signs, structures, and road surface; thus they can’t reliably save you from stopped vehicles.
• Vision‑only and camera+radar systems each have failure modes; lidar is seen by several as the only robust way to detect dark, stationary obstacles at night, but is rare in consumer cars.
• Some believe near‑AGI is needed to safely handle real‑world edge cases; others see clear incremental progress (e.g., Tesla FSD v12) but still far from nap‑in‑the‑car autonomy.

Driving behavior and following distance

• Multiple anecdotes of near‑identical “peek‑a‑boo” events; outcome often hinges on following distance and attention several cars ahead.
• Debate over what constitutes a safe time gap; many ACC implementations are criticized for maintaining too‑short distances and braking abruptly when cut off.

User experiences with BlueCruise, FSD, and ACC

• BlueCruise users describe strict driver‑monitoring (eye tracking, frequent alerts) and limited, mapped‑highway scope; seen by some as well‑designed but still “just” advanced lane‑keeping + ACC.
• FSD users report impressive performance in many scenarios but also occasional sudden, dangerous mistakes (e.g., misjudged turns, odd behavior around stopped or mispositioned cars).
• Phantom braking, emergency‑vehicle handling, and sensor‑removal decisions (e.g., dropping radar) are contentious; some report big improvements, others remain distrustful.

Lighting, visibility, and simple safety measures

• Thread highlights frequent night driving without proper headlights/taillights (daytime running lights only), making stopped cars hard to see.
• Calls for automatic headlights, better cues when lights are off, and wider use of safety triangles, retroreflectors, vests, and flares.

Regulation, testing, and broader policy

• Many want aviation‑style investigation and standardized safety tests for ADAS/FSD (including stopped‑vehicle scenarios), not just for high‑profile systems.
• Disagreement over whether regulators or manufacturers bear primary blame for premature deployment.
• Several argue that focusing on self‑driving misses the larger issue: inherently lethal car‑centric infrastructure and underinvestment in transit and rail.