Because Boeing did nothing to replace a defective part, an airplane crashed killing fifteen

MD-11 crash from November 2025
Click for NTSB report.

Despite previously identifying stress fractures in a part that held the engines to the wing on three different MD-11 airplanes, Boeing did nothing to replace the part, and so fifteen people died when the engine fell off a UPS cargo plane at take-off in November 2025.

The sequence of images to the right, which I have annotated to show the engine breaking free from the wing, comes from the NTSB preliminary investigation report [pdf]. From the article at the link above:

In an update to its ongoing investigation into the crash of UPS Flight 2976, the National Transportation Safety Board [NTSB] said its team found fatigue cracking and overstress failure across much of the bearing race inside the area that attached the plane’s left engine to its wing. NTSB investigators then went back into Boeing service data and confirmed the design of the bearing assembly was consistent with the original design of that part.

[A] Boeing Service Letter dated Feb. 7, 2011 [and found by the NTSB], told operators the company was aware of four previous bearing race failures on three different airplanes. Boeing had seen the fractures of the bearing race, with the parts splitting in two and moving out of place. However, Boeing told operators its review of the bearing failure “would not result in a safety of flight condition.”

Boeing said further regular inspection of MD-11 airplanes would include a look at this bearing assembly, something scheduled for 60-month service intervals. And while Boeing used that service letter to discuss a new bearing assembly configuration, the installation of the original parts “was not prohibited.”

The plane itself had been built by McDonnell-Douglas, prior to its merger with Boeing. Nonetheless, Boeing engineers and managers were aware of this issue and did nothing to inform the owners of this plane so they could take action. In fact, Boeing apparently continued to ship out the original parts to airlines as replacement spares.

This is another example of a serious quality control problem at Boeing, where engineers no longer view serious engineering failures as serious engineering failures.

22 comments

Axiom has delayed the launch of its first space station module to ’28

Axiom's module assembly sequence
Axiom’s module assembly sequence

When Axiom announced in September 2025 that Redwire would be building the solar panels for the first module of its space station, dubbed the PPTM, it also said that module would launch in late 2027, which was a delay of one year from the original launch date of 2026.

That schedule has now apparently been delayed again. In an interview yesterday, the company’s vice president of human spaceflight, former NASA astronaut Peggy Whitson, indicated the launch was now targeting 2028.

Plans call for the initial Axiom Station to be comprised of two modules, the PPTM — short for Payload Power Thermal Module — and a habitat module. The PPTM, which is to be shipped shortly to Houston for final assembly and integration, is slated to be launched in early 2028, with the second module following just months later. From there, Axiom aims to swiftly begin welcoming crew, Peggy Whitson, the company’s vice president of human spaceflight, told me in an interview.

This schedule almost guarantees that the Axiom station will not detach from ISS as quickly as originally intended. PPTM has a large hatch opening connecting it to ISS, allowing for the easy transfer of much of the research racks held on ISS. Before Axiom can become a free-flying station that ISS equipment must be moved, a process that will take time, likely months. To get it done the company will probably have to also attach its second habitation module so that crews can arrive and begin this transfer process.

In other words, Axiom’s schedule margins for getting its station launched, docked to ISS, loaded with ISS equipment, and then separated before ISS retires in 2030 are shrinking. It can ill afford further delays.

Below are my rankings of the five American space stations presently under development. Note that I now consider Axiom and Starlab tied for second.
» Read more

12 comments

Another ESA rendezvous demo mission proposed

The European Space Agency (ESA) and the Luxembourg startup ClearSpace yesterday announced a new demo mission to test autonomous rendezvous and proximity maneuvers, scheduled to launch in 2027.

PRELUDE aims to validate autonomous rendezvous and proximity operations in real flight conditions. The mission will test high-accuracy tracking, navigation and maneuvering using a combination of vision-based and complementary sensors feeding onboard algorithms and autonomous, fault-tolerant guidance, navigation and control (GNC) software. The goal is to demonstrate full freedom of movement and safe, repeatable maneuvers around another spacecraft.

Sounds good, eh? Not so fast. ClearSpace has had a bunch of these missions proposed, and none has yet flown. In 2019 ClearSpace won an ESA contract to de-orbit an old piece of space junk by 2025. In 2023 however that mission was stymied when that space junk, a payload adapter from a 2013 launch of Vega rocket launch, was hit by another piece of space junk.

Both ESA and ClearSpace apparently had difficulties re-designing the mission. In 2024, the ESA forced a major shake-up in ClearSpace’s management and missions, with the established company OHB taking over the startup. Subsequently the mission was redesigned to de-orbit a different defunct satellite, but delayed until 2029.

In 2024 the United Kingdom gave ClearSpace and Japan’s Astroscale a contract to de-orbit two satellites in ’26. It is however not clear at this time whether that mission will launch as planned.

This new PRECLUDE mission is interesting in that it will test the rendezvous and proximity technology that ClearSpace must have for all the other de-orbit missions. In other words, those other missions were never possible, because ClearSpace didn’t have the capability to do them. This new mission appears designed to develop that capability.

I ask: Why wasn’t PRECLUDE scheduled first, in the first place? That it wasn’t reflects very badly on both ClearSpace and the ESA.

0 comments

A small European prototype re-entry capsule survived PSLV launch failure

A small prototype re-entry demonstration capsule, built by the Spanish startup Orbital Paradigm and dubbed the Kestrel Initial Demonstrator (KID), apparently survived for a short period the failure of the third stage of India’s PSLV rocket early this week.

According to an Orbital Paradigm press release, the survival of its little demonstrator came as a surprise. “When we understood that the launch was non-nominal it was a bit of a hit for us,” explained Orbital Paradigm CEO Francesco Cacciatore. … “I think the launch livestream was still ongoing when the team saw that we had 190 seconds of flight data transmitted and received. We needed a few minutes to realize it was real data and not a glitch.”

…“KID was tested beyond its design envelope, and it worked. Separation, power-on, and data transmission, even after reentry, all performed well despite degraded conditions,” explained the company in a 13 January update. “Based on initial analysis, it seems that we achieved 4 out of 5 launch milestones, albeit through an off-nominal profile. The failure to deliver customers’ data prevents us from declaring the mission a success.”

The company considers the mission a failure because it did not get the re-entry data back that it really needed. It says however it is moving forward on a more advanced demonstrator it hopes to launch in 2027. I suspect it will not hire India’s space agency ISRO to launch it.

1 comment

ULA loses another launch contract to SpaceX

The Space Force yesterday announced it has switched rocket companies for its next GPS satellite launch, taking the launch away from ULA and its Vulcan rocket and giving it to SpaceX.

SpaceX could launch the GPS III Space Vehicle 09 (SV09) within the next few weeks, as the satellite was entering the final stages of pre-flight preparations. As part of the swap, United Launch Alliance (ULA) will instead launch the third of the next generation of Global Positioning System satellites. The GPS III Follow-on (GPS IIIF) SV13 satellite was originally scheduled to launch on a Falcon Heavy, but will now fly on Vulcan.

“SV09 and SV13 were traded between ULA and SpaceX to get capability to orbit as soon as possible, for the same reason as the prior swap, which resulted in the last GPS launch in May 2025,” the spokesperson said in a statement. “The trade results in an overall net cost savings to the government and again demonstrates our sustained commitment to moving at speed to deliver combat-credible capabilities on orbit to meet warfighter needs.”

While at first glance it appears ULA has lost nothing, the military’s decision here bodes ill for the company. First, it indicates ULA has been unable to get Vulcan ready on time, forcing the Space Force to look to someone who could.

Second, this is the second time the Pentagon has taken a launch from ULA for these reasons. Increasingly it appears the military is losing patience with ULA’s inability to launch on time. For example, in awarding its most recent set of nine launches, it gave them all to SpaceX, bypassing ULA entirely.

In the past the Space Force tolerated ULA’s delays and high launch cost in order to guarantee the military had more than one launch provider. It now appears it is placing more importance on reliability and cost savings. And as I say, this bodes ill for ULA, which has not done a good job of providing either.

5 comments

Mitsubishi buys space on proposed Starlab space station

Starlab design in 2025
The Starlab design in 2025. Click
for original image.

The Japanese company Mitsubishi has now signed an agreement with the consortium building the large single module Starlab space station that will launch on Starship, reserving part of that station for the company’s own use while also increasing its financial investment in the project.

Starlab Space LLC today announced that Mitsubishi Corporation has reserved and pre-purchased capacity on Starlab’s commercial space station, becoming a foundational customer while simultaneously increasing their investment in the company and joining Starlab’s Board of Directors through representative Issei Shinohara.

The expanded partnership includes acquisition of usage rights for designated payload volume and utilization of on orbit laboratory facilities on Starlab, positioning Mitsubishi to accelerate space-based research opportunities for Japanese institutions. This customer commitment is accompanied by an expanded equity partnership that brings additional investment to support Starlab’s development.

Mitsubishi had in April 2024 already joined the Starlab partnership, though almost no details were announced at that time. Today’s announcement provides those details. It also appears Mitsubishi is bypassing Japan’s space agency JAXA, which in the past always ran such international projects. Instead Mitsubishi will directly “support Japanese space development objectives while contributing to advancements in areas such as life sciences research, advanced materials development, and next-generation manufacturing technologies.” In other words, it is telling JAXA to jump in the lake. It can do this better without that government third party, which by the way has not been very effective in recent years.

This deal continues Starlab’s aggressive momentum in recent months. Though Axiom is still listed above it in my rankings below of all the American space stations under development, I now consider the two essentially tied for second.
» Read more

0 comments

SpaceX launches another 29 Starlink satellites

SpaceX today completed its fifth launch in 2026, placing 29 Starlink satellites into orbit, its Falcon 9 rocket lifting off from Cape Canaveral in Florida.

The first stage competed its 25th flight, landing on a drone ship in the Atlantic.

At this moment the entire 2026 launch race is SpaceX, and SpaceX only. The only other entity to attempt a launch so far in 2026 has been India’s space agency ISRO, and that launch was a failure last night.

0 comments

Space station Starlab gets major new investor

Starlab design in 2025
The Starlab design in 2025. Click
for original image.

The consortium led by Voyager Technologies that is building the Starlab space station announced last week that it has obtained a major new investor with more than a billion dollars in assets.

The investor, Seven Grand Managers, is based in New York. The announcement did not specify how much the firm had committed to the Starlab project, but it was clear from this statement that involved significant funds.

Starlab is being built to be commercially viable from Day One,” said Chris Fahy, founder and chief investment officer, Seven Grand. “Our investment recognizes that commercial infrastructure in the post-ISS era is not speculative, but tangible, bankable and poised for growth. Starlab’s world-class management team and strategic partners are unlocking the beginning of this enormous opportunity.” [emphasis ine]

The highlighted quote suggest Seven Grand was impressed with the Starlab concept, a single very large ready-to-go station launched on Starship. Most of the other stations will involve assembly of multiple modules on multiple launches before they are “ready-to-go.” The only other station launching as a single module, Max Space’s Thunderbird, has only recently entered the race, and is thus far behind.

Starlab had previously raised $383 million in a public stock offering, in addition to the $217.5 million provided by NASA. This new private investment capital further strengthens its future, and suggests the station could get built and launched, even if it fails to win a major station construction contract from NASA.

Below are my rankings of the five American space station projects:
» Read more

1 comment

SpaceX launches NASA’s Pandora exoplanet space telescope

SpaceX today successfully launched a new NASA space telescope, Pandora, its Falcon 9 rocket lifting off from Vandenberg in California.

Pandora is a smallsat focused on studying 20 stars known to have transiting exoplanets. It will look at each repeatedly to draw as much information about the star and the exoplanet as possible. Also deployed were two other NASA smaller astronomy cubesats.

The Falcon 9 first stage completed its 5th flight, landing back at Vandenberg. The two fairing halves completed their first and seventh flights respectively.

At this moment, SpaceX is the only entity to have launched in 2026. This was its fourth launch.

5 comments

Space Force awards nine launch contracts to SpaceX

In announcing its next round of satellite launch awards, the Space Force’s Space Systems Command (SSC) has awarded all nine launches (valued at $739 millon total) to SpaceX, bypassing both Blue Origin and ULA.

SSC awarded the [three] SDA-2 missions to SpaceX for launches projected to begin in [the fourth quarter of fiscal year ’26], and awarded the [two] SDA-3 missions to SpaceX for launches to begin in [the third quarter of fiscal year ’27]. SSC also awarded the [four] NTO-5 launches to SpaceX projected to occur in [the first quarter of fiscal year ’27 and the second quarter of fiscal year ’28]. The total value of these awards is $739M.

It is surprising that SpaceX got all nine contracts. Even though SpaceX charges less than ULA, and Blue Origin’s New Glenn rocket is not yet certified by the Pentagon for operational launches, it has been military policy in recent years to distribute this work to more than one launch provider so as to guarantee a redundancy. ULA exists today for expressly that reason. In the past it would have certainly gotten at least one of these launches.

As for Blue Origin, the Space Force could have awarded it at least one of the later launches in ’27 and ’28, contingent on getting New Glenn certified.

That the Space Force bypassed both companies entirely speaks volumes. It appears it has decided to simply go with the best product now available, and to heck with redundancy.

20 comments
1 24 25 26 27 28 677