Software Bug in Mars Climate Orbiter

1
Software Bug in Mars Climate Orbiter

• Ala Aldin Meriden (Aladdin)
• EE 585-001
• Fall 2006
• In-class Presentation 1
• Dr.James Lumpp

2
Outline

• Introduction
• The problem
• Example
• The results
• The consequences
• Conclusion
• References

3
Introduction

• The name of that mission was Mars Climate Orbiter
  (MCO).
• The MCO spacecraft was lunched by Delta rocket on
  December 11, 1998.
• The MCO was lunched from Cape Canaveral Air
  Station, Florida.
• After 9½ months the spacecraft was to reemerge
  from behind Mars , did not occur.
• The MCO was to have used a technique called
  aerobraking. (Look at the figure next slide).
• The bug occurred because two teams of that
  mission, one was in Colorado and the other was in
  California, were using two different units.

4
Introduction (cont.)
5
The Problem

• The cause of the error was A failure to
  recognize and correct an error transfer of
  information between the MCO spacecraft team in
  Colorado and the mission Navigation team in
  California
• One team used English Unites (e.g. inches, feet
  and pound) while the other used metric units for
  a key spacecraft operation.

6
The Problem (cont.)

• The Angular Momentum Distribution (AMD) file
  contained the output data from the software
  application code, Small Forces, which was using
  the English units.
• On the other hand, MSOP project Software
  Interface Specification was to be in metric
  unites.

7
Example

• 1 inch 2.6 cm
• 12 inch 31.2cm
• 1 cm 0.385 inch
• 12 cm 4.615 inch
• 1 0.385 0.615 inches different

8
The Results

• Therefore, the actual trajectory was lower than
  the expected trajectory, and it was not detected
  because of the mismatch of unites of measurement
  used in the navigation software and the
  propulsion system.

9
The Results (cont.)

• So, they lost communications with MCO, and it is
  presumed to have broken up in the Martian
  atmosphere, or crashed on the planet.
• There were indications that this bug could have
  been picked up, between the times the MCO was
  launched, and just before it was due to enter
  Mars orbit.

10
The Results (cont.)
11
The Consequences

• The total cost of the spacecraft development and
  mission operations of this project was US 235.9
  million.
• Delay of scientific missions
• MSO was designed to help scientists better
  understand the climate history of Mars.
• The MSO was to be the First Martian Weather
  Satellite.
• The MSO carries two science instruments
• The pressure Modulator Infrared Radiometer.
• The Mars Color Imager.
• Concerns over the Mars Polar Lander.
• Concerns over NASA policies.

12
Conclusion

• The root cause for the loss of the Mars Climate
  Orbiter in September 1999, was due to the failure
  to convert the outputted Imperial measurements to
  the required metric measurements, within a
  specific piece of ground software.
• This bug started with a human error, and ended
  with a system failure witch led to the mission
  failure.
• This case study highlights issues of
  interoperability between different components,
  and the difficulty of the verification of a
  system as a whole.

13
References

• Jet Propulsion Laboratory, California Institute
  of Technology, National Aeronautics and Space
  Administration (1999). NASA'S MARS CLIMATE
  ORBITER BELIEED TO BE LOST. http//marsweb.jpl.nas
  a.gov/msp98/news/mco990923.html
• Jet Propulsion Laboratory, California Institute
  of Technology, National Aeronautics and Space
  Administration (1999). Mars Climate Orbiter
  Mission Status. http//marsweb.jpl.nasa.gov/msp98/
  news/mco990924.html
• Isbell, D., Hardin, M. and Underwood, J. (1999).
  MARS CLIMATE ORBITER TEAM FINDS LIKELY CAUSE OF
  LOSS. http//marsweb.jpl.nasa.gov/msp98/news/mco99
  0930.html

14
Questions

• ?