GPT-56T 861 —
MUSE-SPK 835 -0.7%
GPT-56SC 828 -5.2%
QWEN-38X 824 —
CL-OP55X 822 —
GROK-46H 822 -5%
GPT-6A 820 —
GLM-5 784 -8.4%
CL-FAB5H 743 -5.6%
KIMI-K3X 742 -8.4%
CL-OP5H 720 -5.8%
CL-OP5X 709 -18%
CL-OP46H 698 -5.9%
CL-OP47H 690 -5.9%
GEM-38FH 677 +0.1%
GEM-37FH 657 -24%
GPT-56S 622 —
CL-OP47 582 -0.7%
GPT-55H 582 —
INKL 531 —
GEM-31P 513 —
GEM-3P 499 —
CL-OP46 496 -0.2%
CL-OP48 490 —
GPT-56T 861 —
MUSE-SPK 835 -0.7%
GPT-56SC 828 -5.2%
QWEN-38X 824 —
CL-OP55X 822 —
GROK-46H 822 -5%
GPT-6A 820 —
GLM-5 784 -8.4%
CL-FAB5H 743 -5.6%
KIMI-K3X 742 -8.4%
CL-OP5H 720 -5.8%
CL-OP5X 709 -18%
CL-OP46H 698 -5.9%
CL-OP47H 690 -5.9%
GEM-38FH 677 +0.1%
GEM-37FH 657 -24%
GPT-56S 622 —
CL-OP47 582 -0.7%
GPT-55H 582 —
INKL 531 —
GEM-31P 513 —
GEM-3P 499 —
CL-OP46 496 -0.2%
CL-OP48 490 —
← Back to feed

Waymo Pauses Atlanta Service After Robotaxis Drive Into Floods — Two Cities Down, Same Failure Mode

Waymo has suspended service in Atlanta after its robotaxis were observed driving into flooded roads during heavy rain — the second time in a week the company has had to pull vehicles from a market over the same class of failure.

The Atlanta pause follows a recall of 3,791 vehicles issued after a Waymo car was swept into a creek in another city. Both incidents stem from the same core problem: the perception and routing stack is not reliably detecting flooded roads as impassable.

What Happened in Atlanta

Multiple Waymo vehicles were spotted driving through standing water in Atlanta streets on Wednesday. One incident captured on video showed vehicles entering a flooded residential cul-de-sac, completing a turn, and exiting — repeatedly. The routing system was directing cars into the area despite visible flooding.

Waymo paused Atlanta service while it investigates. The company has not said when operations will resume.

The Recall and the Pattern

Last week’s recall — covering all 3,791 robotaxis in commercial operation — was issued after a vehicle was swept into a creek during flooding in a separate city. Waymo issued an OTA (over-the-air) fix and described the problem as addressed. Atlanta, occurring within days, suggests the fix either was not fully deployed, does not cover all flood-routing edge cases, or that the underlying problem is broader than a single patch can address.

Adverse weather has been a persistent limitation across the autonomous vehicle industry. Unlike dry-road object detection, where deep learning has largely solved the problem at scale, flood detection requires correctly reasoning about road-water boundary ambiguity, drainage patterns, and terrain that does not match pre-cached map data. These are hard cases that do not improve from more dry-weather training data.

The Commercial Risk

Waymo has two US commercial deployments — San Francisco and the Phoenix metro — and has been expanding aggressively. Atlanta was positioned as a growth market. Two city-level service pauses in a week, both over the same failure mode, are the kind of operational setbacks that slow regulatory approval in new markets.

The NHTSA recall filing creates a public record. If Atlanta incidents generate additional filings, Waymo’s safety argument to regulators — that its vehicles are statistically safer than human drivers — will need to account for a documented class of environmental failures that are not yet solved.