GAO Report Questions NASA Management of SLS & Orion

Artist’s conception of Space Launch System in Vehicle Assembly Building (Credit: NASA)

NASA management of the Space Launch System, Orion and Exploration Ground Systems (EGS) programs could lead the agency to repeat one of the mistakes that led to the loss of the space shuttle Columbia, according to a new report from the  Government Accountability Office (GAO).

“The approach has dual-hatted positions, with individuals in two programmatic engineering and safety roles also performing oversight of those areas,” the report stated. “These dual roles subject the technical authorities to cost and schedule pressures that potentially impair their independence.

“The Columbia Accident Investigation Board found in 2003 that this type of tenuous balance between programmatic and technical pressures was a contributing factor to that Space Shuttle accident,” the report added.

The GAO analysis also found that NASA has not calculated the cost of flying the system other than for the first flight.

“In May 2014, GAO recommended that NASA baseline the programs’ cost and schedule beyond the first test flight,” the report stated. “NASA has not implemented these recommendations nor does it plan to; hence, it is contractually obligating billions of dollars for capabilities for the second flight and beyond without establishing baselines necessary to measure program performance.”

NASA partially concurred with a GAO recommendation to address the dual hat management issue. However, “NASA did not address the need for the technical authority to be independent from programmatic responsibilities for cost and schedule. GAO continues to believe that this component of the recommendation is critical,” the report stated.

The GAO also recommended that Congress consider requiring NASA to establish separate cost and schedule baselines for the second SLS exploration mission. Congress has not addressed this recommendation yet.

The report’s executive summary follows.

NASA Human Space Exploration
Integration Approach Presents Challenges to Oversight and Independence

United States Government Accountability Office
Report to Congressional Committees
October 2017
GAO-18-28 (PDF)

Why GAO Did This Study

NASA is undertaking a trio of closely related programs to continue human space exploration beyond low-Earth orbit. All three programs (SLS, Orion, and EGS) are working toward a launch readiness date of no earlier than October 2019 for the first test flight. Each program is a complex technical and programmatic endeavor. Because all three programs must work together for launch, NASA must integrate the hardware and software from the separate programs into a working system capable of meeting its goals for deep space exploration.

The House Committee on Appropriations report accompanying H.R. 2578 included a provision for GAO to assess the progress of NASA’s human space exploration programs. This report assesses (1) the benefits and challenges of NASA’s approach for integrating these three programs and (2) the extent to which cross-program risks could affect launch readiness. GAO examined NASA policies, the results of design reviews, risk data, and other program documentation and interviewed NASA and other officials.

What GAO Found

The approach that the National Aeronautics and Space Administration (NASA) is using to integrate its three human spaceflight programs into one system ready for launch offers some benefits, but it also introduces oversight challenges. To manage and integrate the three programs—the Space Launch System (SLS) vehicle; the Orion crew capsule; and supporting ground systems (EGS)—NASA’s Exploration Systems Development (ESD) organization is using a more streamlined approach than has been used with other programs, and officials GAO spoke with believe that this approach provides cost savings and greater efficiency. However, GAO found two key challenges to the approach:

  • The approach makes it difficult to assess progress against cost and schedule baselines. SLS and EGS are baselined only to the first test flight. In May 2014, GAO recommended that NASA baseline the programs’ cost and schedule beyond the first test flight. NASA has not implemented these recommendations nor does it plan to; hence, it is contractually obligating billions of dollars for capabilities for the second flight and beyond without establishing baselines necessary to measure program performance.
  • The approach has dual-hatted positions, with individuals in two programmatic engineering and safety roles also performing oversight of those areas. As the image below shows, this presents an environment of competing interests.


These dual roles subject the technical authorities to cost and schedule pressures that potentially impair their independence. The Columbia Accident Investigation Board found in 2003 that this type of tenuous balance between programmatic and technical pressures was a contributing factor to that Space Shuttle accident.

NASA has lowered its overall cross-program risk posture over the past 2 years, but risk areas—related to software development and verification and validation, which are critical to ensuring the integrated body works as expected—remain. For example, delays and content deferral in Orion and SLS software development continue to affect ground systems software development and could delay launch readiness. GAO will continue to monitor these risks.

What GAO Recommends

Congress should consider directing NASA to establish baselines for SLS and EGS’s missions beyond the first test flight. NASA’s ESD organization should no longer dual-hat officials with programmatic and technical authority responsibilities.

NASA partially concurred with our recommendation and plans to address it in the next year. But NASA did not address the need for the technical authority to be independent from programmatic responsibilities for cost and schedule. GAO continues to believe that this component of the recommendation is critical.

Matter for Congressional Consideration

Congress should consider requiring the NASA Administrator to direct the Exploration Systems Development organization within the Human Exploration and Operations Mission Directorate to establish separate cost and schedule baselines for work required to support SLS and EGS for Exploration Mission 2 and establish separate cost and schedule baselines for each additional capability that encompass all life cycle costs, to include operations and sustainment.

Status: Open

Comments: When we determine what steps the Congress has taken, we will provide updated information.

  • Merisea

    Jerald Mason to Bob Lund: “Take off your engineering hat and put on your management hat.”
    …and seven people died.

  • Andrew_M_Swallow

    There is a term mission in cost estimates
    + (cost of replacement vehicle + third party damage + cost of say 2 year stand down) * probability of accident

    This will make NASA (and BP deep sea drilling) more cautious. Getting the probability of accident down will make a big difference to the cost estimates

  • ThomasLMatula

    Which has nothing to do with this report which focuses on what appears to be spending that is out of control on the SLS because the ones spending the money are also responsible for overseeing the spending. Or are you arguing the engineers should be free to spend taxpayer money without any control or oversight?

  • ThomasLMatula

    Yes, so cautious they will never let the ship leave the dock. But then why build the ship in the first place?

  • Michael Halpern

    Part of the problem is that it was established early on that SLS would not have much of an R&D budget for brand new technology outside of Orion, and that it will be a reconfiguration of shuttle technology, on one hand, it’s proven technology, but on the other, you are taking parts from a LEO vehicle and making them Lunar capable, rather than taking the lessons learned from that vehicle and making something better, all the tweaking existing equipment is probably making it more expensive then new equipment and modernized manufacturing.

  • Michael Halpern

    That is another problem, giving the engineers a blank check, most engineers don’t think about how much something has to cost, only what it does right now, unless they are involved in making the part, and as NASA sources all their parts from various suppliers, no research has been done on modernizing the manufacturing methods from what it was in the 70s leading to a higher price tag

  • Michael Halpern

    Fortunately not all of NASA is so bad with money, unfortunately those ones are only running Commercial Resupply Services and Commercial Crew Services it seems

  • ThomasLMatula

    And what is your evidence NASA would have done better using the old cost-plus contract method for both?

  • Michael Halpern

    The fact that commercial resupply makes financial sense and it is less expensive than any other ISS supply scheme and commercial crew is built on the principles of commercial resupply services

  • ThomasLMatula

    The aerospace firms have made large strides in modernizing their manufacturing methods in the last forty years. Just look at how commercial aircraft and spacecraft are built.

    Or are you referring to NASA contracting methods and their quality control requirements? Now those I could believe are 40 years out of date.

  • ThomasLMatula

    But you indicated above that they weren’t using the money wisely…

  • Michael Halpern

    I am referring to the SLS fuel tank, the new vertical welding tool (verses the rotating welding tool used for the shuttle) is marginally easier to use, costs millions and is slower, and at the end of half a year you still get a ridiculously expensive fuel tank

  • Michael Halpern

    I said the SLS guys weren’t, they aren’t in charge of CRS or commercial crew

    I see how I could have been misinterpreted

  • ThomasLMatula

    Michael, Yes I did.

    I agree they are doing a poor job on the SLS. I also think they could do a better job of managing the ISS. It is the problems with these two flagship programs that give NASA its bad image in terms of human spaceflight.

  • ThomasLMatula

    Yes, but because they wanted to use Shuttle legacy hardware they are stuck with Shuttle legacy technology. Its another reason they really should walk away from the SLS. They would have been much better off just recreating the Saturn V using modern manufacturing technology and materials. It was a good design to start with being based on 30 years of rocket design experience starting with the German A1.

  • You don’t need to keep the cost of the replacement vehicle in that term. The probability of replacing an SLS after a mission–whether it fails or not–is 100%.

  • Vladislaw

    Gosh it took congress 50 years to put those systems in place .. to congressional space state members.. it’s it running perfectly and just how they want it?

  • Michael Halpern

    Yeah I wouldn’t be surprised if the NASA accountants decided to lease 39A to SpaceX until 2034 BECAUSE they felt the SLS was never going to get much use out of it. On one hand, its a historic pad, on the other its almost worse letting it go derelict.

  • Andrew_M_Swallow

    The next SLS is reserved for a different mission and therefore has its own budget. The current mission will normally need completing so the accident is forcing an additional SLS and Orion to be manufactured.

  • Robert G. Oler

    this is a surprise…not

  • Mr Snarky Answer

    Don’t worry no astros will die going to Space if you never finish the vehicle.

  • Jeff2Space

    Actually they have been building “new equipment and modernized manufacturing”. Just look at what they’re doing to build the SLS core stage. This is part of the reason development costs are so high and the schedule keeps slipping. Instead of reusing space shuttle ET manufacturing facilities, fixtures, technologies, and etc., they’re using new fixtures and bleeding edge welding technologies.

  • Michael Halpern

    The vertical welding tool for the core stage is barely an improvement over the rotating welding tool, the bleeding edge of welding is in metal 3d printing not a manually/semi manually operated welder with a camera

  • Michael Halpern

    Also what they could have done is investigate better alloys so they don’t need to have corrugated structure and instead make a big tube via rolling and use that for the core

  • Michael Halpern

    Or you could have a Falcon Heavy or New Glenn on standby

  • Michael Halpern

    One problem is that there is only one supplier for the External tank/ SLS core, and no opportunity was presented for competition, so they can charge almost whatever the hell they want to per core,

  • ThomasLMatula

    Nor will there be any disturbing videos like this for the Ares I on youtube to explain away 🙂

    https://www.youtube.com/watch?v=You6yrw-VL8

  • publiusr

    Not on F-35–on space–sure, as a matter of fact. I want to see Spaceflight with political moxie–same as defense.

  • publiusr

    Now on Ares I–we agree.

  • Paul451

    but because they wanted to use Shuttle legacy hardware they are stuck with Shuttle legacy technology.

    Except they didn’t use legacy hardware. (Nor, as Michael noted, legacy manufacturing.) The tanks are not the same as, or even reasonably modded versions of, the Shuttle’s ETs. They increased the diameter (8 to 10m) and length (40 to 60m), 500,000 to 700,000 gallons (35% increase in volume). They therefore has to increase the thickness of the walls. They changed the alloy they used. They changed the construction/welding method (which it appears Boeing is still struggling to get right.)

    It’s the same with every component. They deliberately chose a design with as little commonality with STS hardware as possible, except superficially.

  • Jeff2Space

    That’s true of single source government contracts in general.

    Commercial Cargo and Commercial Crew are successful in part because NASA has chosen multiple providers for both, so competition exists even after the contracts were awarded. Competition is a good thing.

  • Michael Halpern

    Yeah and almost every major component not made or cannabolised (looking at the SSMEs) directly by NASA or ESA has at most 2 suppliers and that’s the SRBs,

  • Michael Halpern

    There is a reason why Elon never puts the SLS on his charts, it will never get a respectable cadence or launch history practically by design, 3 or 4 launches in a 6-8 year period is most likely

  • Michael Halpern

    Commercial Resupply and Commercial Cargo also encourage further competitive development outside of the scope of getting to and from the ISS, many of the companies that produced the shuttle parts and are producing the SLS parts have little desire or inclination to make anything for anyone other than NASA, which is why for many parts there is only a single supplier that is still operating, SLS is the last NASA “jobs program” project, and unlike previous programs, it isn’t doing anything special, not for the cost, after SLS they won’t be able to be so wasteful.

  • Andrew_M_Swallow

    The SLS is a political rocket. The R100 airship was forcibly cancelled when its government built rival the R101 crashed. Avoid the Falcon Heavy and New Glenn suffering a similar fate.

  • Michael Halpern

    Different political environment, and both the r100 and the r101 were similar airships designed by Hughes, and they didn’t exactly have a black box to tell them quickly what went wrong so they could fix it. Falcon Heavy, New Glenn and SLS are 3 dissimilar rockets with abort and escape systems and sensor and recording equipment that Hughes could only dream of, and the launch industry basically works by “we will keep you from launching until you find and rectify what went wrong then you can go ahead as much as you want”

  • Michael Halpern

    Also if the mission is manned (not talking cargo) and you either need to get crew to the moon or rescue the crew any auto pilot capable crew carrying spacecraft capable of meeting up with the stranded crew will do

  • Andrew_M_Swallow

    Sold correctly using a commercial spacecraft as a lifeboat may be possible. They can be docked for 6 months. By refuelling ULA has worked out a way of getting spacecraft to lunar orbit.