Professionalism/Mars Climate Orbiter

Mission edit

 
Mars Surveyor 98 Mission Patch
 
Mars Climate Orbiter during assembly

Mars Climate Orbiter was a NASA Jet Propulsion Laboratory mission to analyze the weather and climate of Mars. Spacecraft design and construction were contracted through Lockheed Martin. Mars Climate Orbiter was part of the Mars Surveyor '98 program along with the Mars Polar Lander. Its mission objectives included:

  • Investigating the distribution and composition of surface materials such as dust, carbon dioxide, and water
  • Studying Martian weather patterns with focus on dust storms
  • Determining whether the ancient Martian climate was temperate
  • Relay communications from the Mars Polar Lander[1].

Mars Climate Orbiter launched on a Delta II rocket from Cape Canaveral Air Force Station on 11 December 1998. It was to enter a highly elliptical Martian orbit in September 1999. For two months, it was supposed to slowly circularize its orbit by skimming the upper Martian atmosphere in a process called aerobraking, reaching its final orbit about two weeks before the arrival of the Mars Polar Lander in November 1999[1].

Loss of Communication edit

When Mars Climate Orbiter arrived at Mars on 23 September 1999, mission control expected to lose contact at 09:06 UTC as the craft passed behind Mars. Contact was lost two minutes earlier, indicating that its trajectory was different than planned. Mars Climate Orbiter did not exit the Martian shadow as expected at 09:27 UTC, and contact with the orbiter was never reestablished[2]. NASA declared the mission a loss the next day and launched an investigation.[3]

Background edit

Faster, Better, Cheaper edit

 
NASA Administrator Daniel Goldin

The Mars Climate Orbiter mission was launched under the 1990s NASA mantra of "Faster, Better Cheaper." This policy was instituted by the ninth NASA administrator, Daniel Goldin, when he assumed the office in April 1992[4]. This philosophy came in the wake of slow, expensive missions such as the Mars Observer, launched in 1992. Mars Observer spent eight years in development before launch[5] and was lost just before Mars orbital insertion at a total cost of $813 million[6]. "Faster, Better, Cheaper" meant drastically reducing development time and mission cost for robotic spacecraft. Goldin encouraged frequent, low-cost launches. He believed NASA must accept some robotic spacecraft losses as inevitable in order to achieve the best scientific results[7]. Goldin summarized "Faster, Better, Cheaper" in 1992: "By building them assembly line style, we can launch lots of them, so if we lose a few due to the riskier nature of high technology, it won’t be the scientific disaster or blow to national prestige that it is when you pile everything on one probe and launch it every ten years"[8]. Mars Climate Orbiter was significantly faster and cheaper than previous missions. It spent only four years in development[9] and the combined cost of Mars Climate Orbiter and Mars Polar Lander was $328 million[10].

One Team Fits All edit

One of NASA’s cost-cutting strategies under “Faster, Better, Cheaper” involved recycling engineering teams for missions. The same team was used for three concurrent Mars Exploration Program missions, including the Mars Climate Orbiter[11]. Subject matter experts were hired by the hour to perform design reviews, instead of being hired full time. This method cut costs and quickly developed experience for the engineers, but led to overworked team members who were not able to gain in-depth systems knowledge of the Mars Climate Orbiter. The team struggled to recognize the problems with the spacecraft during its nine-month trip to Mars, and even when the trajectory problem was discovered, the team was unprepared to solve it[12].

Investigation edit

The investigation board found mismatched units to be the root cause of the mission failure.

AMD Events edit

Mars Climate Orbiter, like many spacecraft, used three-axis reaction wheels to maintain constant attitude. Over time, the solar wind and other perturbations can induce a spin in spacecraft. Whenever the onboard computers detected a spin, Mars Climate Orbiter would spin one or more reaction wheels in the opposite direction to counteract it. These reaction wheels could eventually reach a very high rate of spin and become "saturated" with angular momentum. To slow the wheels while maintaining attitude, Mars Climate Orbiter had to fire its thrusters; this process is called an Angular Momentum Desaturation (AMD) event. Mars Climate Orbiter's mission required an unusually high number of AMD events - about ten times more than expected. This was likely due to its asymmetry; the large solar panel on one side would catch the solar wind and induce a strong spin[2].

Lockheed Martin and SM_FORCES edit

Each AMD event causes a small change in a spacecraft's trajectory. Mars Climate Orbiter sent data about every AMD event back to Earth so that course adjustments could be planned. The ground computers calculating trajectory used a script called SM_FORCES written by Lockheed Martin, which listed thruster impulse values in English units (pound seconds) instead of the metric units (Newton seconds) used by the rest of the program. These computers therefore underestimated the effect of each AMD event by a factor of 4.45[2].

Trajectory Error edit

 
Mars Climate Orbiter's actual and planned trajectories

Because each AMD event was not fully corrected for, Mars Climate Orbiter was off course during its final approach. Its initial orbit reached a minimum altitude of 57 kilometers instead of the planned 226 kilometers. This was well below the estimated threshold for survival at 80 kilometers; Mars Climate Orbiter likely disintegrated in the atmosphere[2].

TCM-5 Discussion edit

Before the orbital insertion, data from the orbiter indicated it was at an altitude of 110 kilometers, just 30 kilometers above the minimum allowable altitude. Navigation engineers discussed executing an emergency maneuver, TCM-5, to increase the spacecraft altitude to a safer level[11]. The team decided against the maneuver because it would have delayed orbital insertion and interrupted communications with the Mars Polar Lander, possibly jeopardizing the Lander’s mission. Procedures for the maneuver had also not been fully developed in advance. The mishap investigation board later reported that performing TCM-5 might have saved the spacecraft[12].

Organizational Errors edit

Lack of Communication edit

Lockheed Martin ignored the published NASA standard to use metric units for all calculations[13]. No group or individual from either side checked the unit specifications and units used during development. A press release cited this communication failure as the root cause of the Mars Climate Orbiter but said that adequate review from either side could have prevented failure[14].

When NASA engineers realized that the spacecraft was off course they planned a TCM-5 maneuver. This maneuver may have saved the spacecraft but it was never performed due to communication and decision making failures[15]. This could have been caused by the management ideology for engineers to prove something wrong, as opposed to a safer approach to prove something is right. NASA staff knew something was wrong but they did not act because they were uncertain.

Lack of Review edit

"People sometimes make errors. The problem here was not the error, it was the failure of NASA's systems engineering, and the checks and balances in our processes to detect the error. That's why we lost the spacecraft," Dr. Edward Weiler, NASA's Associate Administrator for Space Science, stated in a public release regarding the incident[13]. Adequate testing and reporting systems can uncover errors before they become disasters. Chief investigator Arthur Stephenson stated at a press conference, "Had we done end-to-end testing, we believe this error would have been caught"[15]. Inadequate testing due to the "Better, Faster, Cheaper" mindset resulted in spacecraft failure.

Related Cases edit

These organizational failures are a common cause for disaster within an organization. Problems in management can have a devastating impact.

Melbourne-Voyager Collision edit

On February 10, 1964, HMAS Voyager, a destroyer, and HMAS Melbourne, a carrier, were performing flying exercises while using only operational and navigation lights. The Voyager was ordered to move its position from starboard to port[16]. HMAS Voyager altered their course and leadership on the Melbourne assumed that the Voyager was performing a "fishtailing" maneuver[17]. Instead the Voyager crossed without fishtailing, resulting in a collision that sank the Voyager and killed 82 officers and sailors[16]. The communication failure between the ships resulted in tragedy.

Deepwater Horizon Explosion edit

On April 20, 2010 a BP oil rig, Deepwater Horizon, experienced a catastrophe resulting in an explosion killing 11 people[18]. Before the explosion the well had a history of flammable gas "kicks" in which volatile gas escapes through the well and has the potential to react explosively on the rig[18]. Despite knowing the risks and in violation of standard procedure, they replaced the dense core, the last line of defense against an explosion, with a lighter core without a fully-installed plug to get the rig active faster[18]. This mistake was coupled with mechanical failures of well equipment, the first lines of defense, that could've been fixed with more testing and regulation[19]. The push to deliver "faster" caused safety oversight.

Decision Traps edit

Status Quo Bias edit

The decisions of the orbiter's team are symptomatic of status quo bias. Status quo bias is the tendency to make decisions that perpetuate the status quo, and any change from that baseline is perceived as a loss[20]. By performing the TCM-5 maneuver, the team would have opened themselves to criticism or regret. The team feared veering from the ideal timeline because of the Mars Polar Lander, which failed nonetheless due to unrelated issues[11]. Instead of focusing on the risk to the orbiter, which was apparent in the data, they justified inaction by citing possible risk to the Lander. Instead of making an informed decision, the team was influenced by the fear of disrupting the status quo. Even though they had the information necessary to make a decision and perform TCM-5, they ultimately gave more weight to the risk of action than inaction.

Overconfidence Bias edit

The orbiter's team failed to recognize the trajectory issue until the spacecraft was lost. Throughout the spaceflight, they received anomalous data about the orbiter's position. Before insertion, the orbiter was over 100 km below the projected altitude. To a layperson, this would be cause for concern, but the Mars Climate Orbiter's team was overconfident in the spacecraft’s design and subsystems. They assumed that the data was in error and that the spacecraft's position could not be wrong[11]. According to the mishap review board, they were under the perception that orbiting Mars is “routine”[12]. This is synonymous with overconfidence bias in which a person is overconfident in their abilities and judgment[21]. This is described in an internal Jet Propulsion Laboratory memo after the orbiter failure: "There might have been some overconfidence, inadequate robustness in our processes, designs, or operations, inadequate modeling and simulation of the operations, and failure to heed early warnings"[22].

Anchoring Trap edit

The engineers working on the climate orbiter mission were also simultaneously working on the Mars Polar Lander and Mars Global Surveyor. The team did not adequately research the orbiter's systems and had an incomplete understanding of its design. Instead, they relied heavily on their knowledge of the Mars Global Surveyor, which successfully entered orbital insertion 2 years before the Mars Climate Orbiter. This mentality is characteristic of the anchoring trap. This decision trap is caused when the decision-maker relies so heavily on previous knowledge and experience that they fail to recognize how the current situation is different[20]. This leads to decisions that are not appropriate for the current problem. Steve Lilley of NASA’s Safety Center explains how the orbiter's team fell into the anchoring trap: “With oversubscribed team members, minimal training and an incomplete knowledge of Mars Climate Orbiter's design, the operations navigation team relied on their intimate knowledge of the Mars Surveyor to operate Mars Climate Orbiter. Assumptions based on this prior knowledge contributed to mission failure”[11].

Conclusion edit

Three important lessons can be learned from the Mars Climate Orbiter mission and applied to any collaborative engineering. First, expectations for work performed must be understood by all parties. The failure's root cause was Lockheed Martin's ignorance of NASA's unit requirements. Second, engineers must be adequately trained for the task and be held accountable for the result. Mars Climate Orbiter's team was not familiar with the craft and did not recognize or correct the error during flight. Finally, attention to detail is paramount for engineering projects. This loss might have been prevented if NASA and Lockheed Martin engineers had correctly verified their product. "Faster, Better, Cheaper" was not effective here because the Mars Climate Orbiter was made faster and cheaper at the expense of reliability. Further study on this topic should determine how the Mars Climate Orbiter's failure influenced subsequent spacecraft designs and personnel organization. NASA's recent robotic mission successes could signify a better balance between cost and reliability.

References edit

  1. a b National Aeronautics and Space Administration (September 1999). "Mars Climate Orbiter Arrival Press Kit". Press release. http://www.jpl.nasa.gov/news/press_kits/mcoarrivehq.pdf. 
  2. a b c d Mars Climate Orbiter Mishap Investigation Board (10 November 1999), Phase I Report (PDF)
  3. NASA Jet Propulsion Laboratory (September 24 1999). "Mars Climate Orbiter Mission Status". Press release. http://mars.jpl.nasa.gov/msp98/news/mco990924.html. 
  4. Elvia Thompson; Jennifer Davis (November 4 2009). Daniel Saul Goldin (Report). National Aeronautics and Space Administration. http://www.hq.nasa.gov/office/pao/History/dan_goldin.html. 
  5. Gnoffo, Anthony (September 1 1993). "Mars Observer's Creators Take Its Loss Personally At Martin Marietta, 500 Workers Spent 10 Years On The Nasa Project. Now They're Searching For Answers.". Philidelphia Inquirer (East Winsor, NJ). http://articles.philly.com/1993-09-01/business/25986978_1_mars-observer-martin-marietta-astro-space-red-planet. 
  6. Mars Observer (Report). NASA National Space Science Data Center. November 4 2009. http://nssdc.gsfc.nasa.gov/nmc/spacecraftDisplay.do?id=1992-063A. 
  7. McCurdy, Howard (November 14 2001). Faster, Better, Cheaper: Low-Cost Innovation in the U.S. Space Program. Johns Hopkins University Press. {{cite book}}: Check date values in: |date= (help)
  8. Goldin, Daniel (October 21 1992). From Entropy to Equilibrium: a New State of Nature for NASA. National Aeronautics and Space Administration. {{cite book}}: Check date values in: |date= (help)
  9. Edward A Euler; Steven D Jolly; H H Curtis (February 4 2001). The failures of the Mars Climate Orbiter and Mars Polar Lander: a perspective from the people involved (Report). American Astronautical Society. http://web.mit.edu/16.070/www/readings/Failures_MCO_MPL.pdf. 
  10. NASA Jet Propulsion Laboratory, Mars Climate Orbiter Fact Sheet
  11. a b c d e Lilley, Steve (2009). [nsc.nasa.gov/SFCS/SystemFailureCaseStudyFile/Download/469 Lost in Translation] (Report). NASA. nsc.nasa.gov/SFCS/SystemFailureCaseStudyFile/Download/469. 
  12. a b c Mars Climate Orbiter Mishap Investigation Board (Report). NASA. November 10, 1999. ftp://ftp.hq.nasa.gov/pub/pao/reports/1999/MCO_report.pdf. 
  13. a b NASA Jet Propulsion Laboratory (September 30 1999). "MARS CLIMATE ORBITER TEAM FINDS LIKELY CAUSE OF LOSS". Press release. http://mars.jpl.nasa.gov/msp98/news/mco990930.html. 
  14. NASA Jet Propulsion Laboratory (November 10 1999). "MARS CLIMATE ORBITER FAILURE BOARD RELEASES REPORT, NUMEROUS NASA ACTIONS UNDERWAY IN RESPONSE". Press release. http://mars.jpl.nasa.gov/msp98/news/mco991110.html. 
  15. a b Oberg, James (December 1 1999). Why the Mars Probe went off course (Report). IEEE Spectrum. http://nssdc.gsfc.nasa.gov/nmc/spacecraftDisplay.do?id=1992-063A. 
  16. a b Ferry, David (April 23 2014). Melbourne-Voyager: causes and inquiries (Report). Australian Naval Institute. http://navalinstitute.com.au/melbourne-voyager-causes-and-inquiries/. 
  17. Melbourne–Voyager collision. World Heritage Encyclopedia.
  18. a b c Eley, Tom (May 14 2010). What caused the explosion on the Deepwater Horizon? (Report). World Socialist Web Site. https://www.wsws.org/en/articles/2010/05/spil-m14.html. 
  19. Deepwater Horizon Accident Investigation Report (Report). BP. September 8 2010. http://www.bp.com/content/dam/bp/pdf/gulf-of-mexico/Deepwater_Horizon_Accident_Investigation_Report.pdf. 
  20. a b Hammond, Keeney, Raiffa, John, Ralph, Howard (2006). The Hidden Traps in Decision Making (Report). Harvard Business Review. https://hbr.org/2006/01/the-hidden-traps-in-decision-making. 
  21. Dobelli, Rolf (June 11, 2013). The Overconfidence Effect (Report). Psychology Today. https://www.psychologytoday.com/blog/the-art-thinking-clearly/201306/the-overconfidence-effect. 
  22. Oberg, James (December 1999). Why the Mars Probe Went Off Course (Report). Spectrum Magazine. http://faculty.up.edu/lulay/failure/MarsClimateOrbit.pdf.