FAA completes investigation into SpaceX's second fiery Starship test

SpaceX starship fully stacked

Image Credits: SpaceX (opens in a new window)

The Federal Aviation Administration has concluded its review of SpaceX’s investigation of the second Starship launch in November, with the regulator saying Monday that it accepted the “root causes and 17 corrective actions” identified by the company.

While this means the investigation is now closed, SpaceX must implement all the corrective actions and apply for a modified launch license before it can fly Starship again.

“The FAA is evaluating SpaceX’s license modification request and expects SpaceX to submit additional required information before a final determination can be made,” the regulator said in a statement Monday.

SpaceX’s second orbital flight test of the nearly 400-foot-tall Starship rocket in November went farther than the first test by a huge margin: all 33 of the Raptor engines on the Super Heavy booster powered on successfully and none failed in the course of the nearly three-minute ascent burn. The company also pulled off a spectacularly difficult “hot-stage separation” for the first time, wherein the Starship upper stage lit up to push away from the booster.

But like the first test, which occurred last April, it ended in a fiery explosion mid-air. When any rocket launch encounters catastrophic issues during flight, the FAA opens what’s known as a “mishap investigation” — that’s what’s happened here. These are led by the company and reviewed by regulators, who must sign off on the analysis and ensure the corrective actions are actually implemented. After the first test, the FAA directed SpaceX to complete 63 corrective actions.

Per the FAA, the company ultimately identified seven changes to the booster and 10 to the Starship vehicle:

“Seven booster corrective actions were identified, including redesigns of vehicle hardware to increase tank filtration and reduce slosh, updated thrust vector control system modelling, reevaluation of engine analyses based on OFT-2 data, and updated engine control algorithms. Ten Starship corrective actions were identified, including hardware redesigns to increase robustness and reduce complexity, hardware changes to reduce leaks, operational changes eliminating pre-second engine cutoff propellant dumps, flammability analysis updates, installation of additional fire protection, creation of analytical guidance, performance of transient load analysis, and modeling updates.”

In an update also published Monday, SpaceX demystified some of these corrective actions in a rare post-launch postmortem, with the company noting that the likely root cause of the booster’s “rapid unscheduled disassembly” — aerospace lingo for “explosion” — was a filter blockage in one of the Raptor engines. This engine failed “energetically,” the company said, which led to a total loss of the booster.

The Starship upper stage met its own end around seven minutes into the flight, when a propellant leak started a fire that severed communication between the spacecraft’s onboard computers. This led to an engine shutdown, which then triggered the autonomous flight termination system to destroy the vehicle.

Some of these hardware changes and design upgrades have already been introduced to the versions of the vehicle that will fly for the third orbital flight test, SpaceX said in the update. Company CEO Elon Musk said during a conversation on his social media platform, X, that mission could take place as soon as the second week of March.

“More Starships are ready to fly, putting flight hardware in a flight environment to learn as quickly as possible. Recursive improvement is essential as we work to build a fully reusable launch system capable of carrying satellites, payloads, crew, and cargo to a variety of orbits and Earth, lunar, or Martian landing sites,” SpaceX said in the update.

Ultimately, SpaceX is aiming to complete the full mission profile, which includes the Super Heavy booster completing a “boost back burn” and executing a controlled splashdown in the Gulf of Mexico. Starship would then fly halfway around the world and land in the Pacific Ocean.

Waymo's robotaxis under investigation after crashes and traffic mishaps

waymo driverless jaguar i pace

Image Credits: Kirsten Korosec

Waymo’s autonomous vehicle software is under investigation after federal regulators received 22 reports of the robotaxis crashing or potentially violating traffic safety laws by driving in the wrong lane or into construction zones.

The National Highway Traffic Safety Administration’s Office of Defects Investigation (ODI) says the probe is intended to evaluate the software and its ability to avoid collisions with stationary objects, and how well it detects and responds to “traffic safety control devices” like cones. The investigation is designated as a “preliminary evaluation,” which the ODI tries to resolve within eight months.

“NHTSA plays a very important role in road safety and we will continue to work with them as part of our mission to become the world’s most trusted driver,” Waymo said in a statement to TechCrunch.

It’s the second investigation into autonomous vehicles that ODI has publicly announced in the last two days. On Monday, ODI opened a probe into Amazon-backed Zoox’s AVs after receiving two reports of the company’s autonomous-equipped Toyota Highlanders being rear-ended by motorcycles after the SUVs unexpectedly triggered the brakes.

The investigation into Waymo’s software also comes just three months after Waymo issued its first-ever recall of its autonomous software, after two of its vehicles crashed into the same towed pickup truck in Phoenix, Arizona.

The company’s robotaxis have had enough trouble with construction sites that videos of these mishaps have regularly gone viral. Some of these are cited in ODI’s report, like when one of Waymo’s robotaxis drove off a paved roadway into a construction zone in Phoenix last October and sustained underbody damage.

There are more typical fender-benders cited, too. In San Francisco, California last year, one of Waymo’s AVs was waiting to merge into traffic when it decided to re-route. As a result, one of its exterior sensors clipped an SUV. In a May incident in San Francisco, a Waymo AV ran into the bumper of a parked car while trying to execute a “pullover maneuver.”

Many of the crashes cited in ODI’s report, though, tend to cite more mundane encounters.

There are multiple examples where Waymo’s robotaxis had trouble navigating automatic gates at parking complexes. At times, Waymo’s AVs crashed into the gates. In a February incident in Arizona, the Waymo AV encountered a closed gate and, when turning to leave the area, backed into parking spikes and popped its tire. In another from November, a Waymo AV crashed into a chain separating part of a parking lot.

While these aren’t life-and-death scenarios, they help illustrate the hard — and hard-to-predict — corner cases that stand in the way of truly autonomous vehicles.